An adaptive pumping control method and pumping system for groundwater monitoring

Abstract

The present application relates to the technical field of groundwater pollution monitoring, and provides a self-adaptive pumping control method and pumping system for groundwater monitoring, which comprises: trial pumping with a trial flow rate and calculating aquifer water yield capacity parameters, and self-adaptive configuration of pumping parameters according to the grading results; pumping with a rapid update flow rate to update the water body in the well, and switching to low disturbance sampling flow rate low disturbance pumping when approaching the formation pressure balance point according to the water level change trend parameters and pump working condition data; collecting water level recovery data after stopping pumping to determine the stable recovery window and perform curve fitting, and outputting the corresponding confidence aquifer hydraulic parameters to adjust the next cycle pumping parameters. The present application realizes self-adaptive optimization of pumping strategy and synchronous acquisition of hydraulic parameters, and improves the reliability and representativeness of monitoring data.

Description

Technical Field

[0001] This invention belongs to the field of water pollution monitoring technology, and provides an adaptive pumping control method and pumping system for groundwater monitoring. Background Technology

[0002] Shallow groundwater contamination is characterized by its insidious nature, sudden onset, and easy attenuation of early signals. After pollutants infiltrate the groundwater from the surface or leak through pipelines, pools, or other structures, they typically first affect the vicinity of the water table or the upper part of the aquifer, migrating and spreading along the flow direction under the influence of groundwater dynamics. Because early contamination often exhibits strong localization, short duration, and small spatial extent, if the water intake layer of the monitoring well is not controlled, or if water from different depths within the well mixes, shallow contamination signals are easily diluted or masked, leading to a delay in the identification of contamination events.

[0003] In shallow groundwater pollution monitoring, to obtain representative formation water samples and identify early pollution signals, it is usually necessary to pump and replace the water stagnant in the monitoring well. Existing technologies often employ fixed-flow pumping or simple low-flow sampling, i.e., pumping water to the surface for measurement within a preset cycle. However, in aquifers with low permeability or low yield, excessive flow can lead to a sudden drop in water level, while insufficient flow results in prolonged replacement time and difficulty in achieving steady-state criteria. Furthermore, existing monitoring processes typically focus only on water quality parameters, making it difficult to dynamically optimize subsequent monitoring strategies based on actual hydrogeological conditions. This limits the reliability of online monitoring data and its adaptability to complex aquifers. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention provides an adaptive pumping control method and pumping system for groundwater monitoring, which can adaptively configure pumping parameters based on the aquifer's water discharge capacity and simultaneously deduce hydraulic parameters during the monitoring process.

[0005] The technical solution of this invention includes conducting a limited-time trial pumping at a test flow rate, collecting initial static water level and water level response data during the pumping process, calculating aquifer water production capacity parameters, classifying the aquifer water production capacity according to the aquifer water production capacity parameters, and configuring corresponding pumping parameters based on the classification results. The pumping parameters include a rapid update flow rate, a low-disturbance sampling flow rate, an upper limit for drawdown rate, and a maximum drawdown, wherein the rapid update flow rate is greater than the low-disturbance sampling flow rate, and the drawdown rate corresponding to the low-disturbance sampling flow rate is not greater than the upper limit for drawdown rate. Pumping is performed at a rapid update flow rate to update the static water column in the well casing, the water retained in the shallow intake isolation section, and the residual water in the measurement front-end passage. Real-time data is collected during the pumping process. Water level data and pump operating data are collected. Based on real-time water level data, water level change trend parameters are calculated. Combined with pump operating data, it is determined whether the well casing is approaching the formation pressure equilibrium point. When approaching the formation pressure equilibrium point, the pumping flow rate is smoothly switched from rapid update flow rate to low-disturbance sampling flow rate. Low-disturbance pumping is performed using the configured low-disturbance sampling flow rate to obtain representative formation water from the target stratum. Pumping stops when the maximum drawdown is reached. Water level recovery data is continuously collected throughout the entire period. A stable recovery window is determined based on the water level recovery data. Curve fitting is performed based on the water level recovery data corresponding to the stable recovery window. Based on the curve fitting results, the hydraulic parameters of the aquifer with corresponding confidence levels are output, thereby adjusting the pumping parameters for the next monitoring cycle.

[0006] Optionally, the water level change trend parameters include the derivative of the drawdown curve and / or the drawdown per unit pumping volume increment; the method for determining the proximity to the formation pressure equilibrium point includes: when the derivative of the drawdown curve is lower than a preset derivative threshold and the duration is not less than a preset duration, and / or the drawdown per unit pumping volume increment is lower than a preset volume threshold, and at the same time the pump operating data meets the stability constraints, it is determined that the proximity to the formation pressure equilibrium point is reached.

[0007] Optionally, the derivative of the drawdown curve is obtained through sliding window regression or smoothing difference, and the stability constraints include that the flow fluctuation coefficient does not exceed a threshold and that the correlation characteristics between pump current and flow rate are within the normal operating range.

[0008] Optionally, the method for determining the stable recovery window based on water level recovery data includes: identifying a time interval satisfying the quasi-steady flow assumption of well flow theory from the full-time water level recovery data according to a preset recovery criterion; the recovery criterion includes at least one of the following: the error of water level recovery to the reference water level does not exceed 0.02m to 0.10m; the absolute value of the water level change rate does not exceed 0.1cm / min to 0.5cm / min and the duration is not less than 120s to 600s; the real-time fitting residual of the water level recovery curve is less than a preset threshold and multiple consecutive sampling points remain stable.

[0009] Optionally, the curve fitting method includes: using at least one selected from the unit gradient equation, the Cooper-Jacob recovery method, or the exponential recovery model to perform curve fitting on the water level recovery data within the stable recovery window.

[0010] Optionally, the curve fitting also includes validity verification, and the validity verification criteria include: the regression determination coefficient of the curve is not less than a preset value, and the root mean square error is not greater than a preset value; the water level recovery curve in the fitting interval is monotonically increasing and has no abnormal fluctuations; there is no cavitation, blockage or abnormal valve position during the pumping process.

[0011] Optionally, the step of outputting aquifer hydraulic parameters with corresponding confidence levels based on the curve fitting results includes: outputting high-confidence conduction coefficients and permeability coefficients when the fitting results meet all validity verification criteria; outputting medium-confidence conduction coefficients and permeability coefficients when the fitting results partially meet the validity verification criteria; and outputting low-confidence equivalent hydraulic parameters when the fitting results do not meet the validity verification criteria.

[0012] Optionally, adjusting the pumping parameters for the next monitoring cycle based on the hydraulic parameters of the aquifer with corresponding confidence levels includes: for high-confidence hydraulic parameters, the pumping parameters can be fully adjusted, with an adjustment range not exceeding ±30%; for medium-confidence hydraulic parameters, only the low-disturbance sampling flow rate and pump shutdown recovery time are slightly adjusted, with an adjustment range not exceeding ±15%; for low-confidence hydraulic parameters, no pumping parameters are adjusted, the effective setpoint of the previous cycle remains unchanged, and encrypted monitoring is triggered.

[0013] Optionally, safety constraints are also included, which include: real-time monitoring of water level, flow rate and pump current throughout the pumping process; immediately executing speed reduction, pump stop or alarm operation when the maximum pumping duration, minimum allowable water level or maximum pump current threshold is triggered; and executing venting or short-term pump stop protection actions when cavitation, dry running or blockage risk is detected.

[0014] Furthermore, this invention also provides a groundwater monitoring pumping system using the adaptive pumping control method for groundwater monitoring as described above. The system includes a pumping update unit, a water level measurement unit, a process sensing unit, a valve group switching unit, a measurement unit, a discharge / collection and anti-backflow unit, and a controller. The controller is communicatively connected to the pumping update unit, water level measurement unit, process sensing unit, valve group switching unit, and measurement unit. The pumping update unit's water path is connected to the valve group switching unit, and the valve group switching unit is connected to the measurement unit and the external discharge pipeline. The pumping update unit receives controller commands and performs trial pumping, rapid flow update pumping, smooth flow switching, and low-disturbance sampling flow low-disturbance pumping actions, stopping operation when the maximum drawdown is reached. The water level measurement unit collects initial static water level, water level response data during the pumping process, and water... The system collects water level data, including water level rise data throughout the entire period after pump shutdown, and transmits the data to the controller. The process sensing unit collects pump operating data and uploads it to the controller. The valve group switching unit coordinates water flow switching to ensure water delivery and pipeline water renewal. The measurement unit receives representative formation water from the target stratum. The controller calculates aquifer discharge capacity parameters based on the collected data, classifies aquifer discharge capacity, and adaptively configures rapid update flow rate, low-disturbance sampling flow rate, upper limit of drawdown rate, and maximum drawdown. It also calculates water level change trend parameters based on water level data, determines the formation pressure balance point based on pump operating data, and controls flow rate switching. Furthermore, it determines a stable recovery window based on water level rise data, performs curve fitting on the corresponding data, outputs aquifer hydraulic parameters with corresponding confidence levels, and adjusts the pumping parameters for the next monitoring cycle accordingly.

[0015] The technical solution provided by this invention has the following advantages compared with the prior art: By calculating the aquifer's water production capacity parameters through trial pumping and configuring pumping parameters in stages, the problem of fixed-flow pumping being unable to adapt to different aquifer conditions in existing technologies was solved. After rapidly updating the water volume in the well casing using a rapid-update flow rate, the formation pressure balance point was determined based on water level change trend parameters and pump operating data. This allowed for a smooth switch to low-disturbance sampling flow rate for low-disturbance pumping, avoiding the risk of dry burning in low-permeability aquifers and the defects of insufficient renewal in high-permeability aquifers. By curve fitting of water level recovery data after pump shutdown, the hydraulic parameters of the aquifer were output, and the pumping parameters for the next cycle were adjusted accordingly. This achieved dynamic optimization of the monitoring strategy, significantly improving the reliability of monitoring data and adaptability to complex aquifers.

[0016] Other advantages, objectives and features of the present invention will become apparent in part from the following description, and in part from those skilled in the art through study and practice of the invention. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a schematic diagram of the structure of an online monitoring system for shallow groundwater pollution according to an embodiment of the present invention; Figure 2 This is a schematic diagram of shallow water intake and isolation in an embodiment of the present invention; Figure 3 This is a schematic diagram of the valve group switching, discharge collection, and backflow prevention structure according to an embodiment of the present invention; Figure 4 This is a flowchart of an online monitoring method for shallow groundwater pollution according to an embodiment of the present invention; Figure 5 This is a schematic diagram of data quality marking and anomaly diagnosis according to an embodiment of the present invention; Figure 6 This is a schematic diagram of regional collaborative monitoring driven by virtual sensing and remote sensing according to an embodiment of the present invention; Figure 7 This is a schematic diagram illustrating the fully integrated autonomous calibration and fluid processing of an embodiment of the present invention; Figure 8 This is a schematic diagram of the abnormal scenario matrix and handling process according to an embodiment of the present invention.

[0019] Figure label: 1. Shallow well water intake and isolation unit; 2. Pumping and renewal unit; 3. Valve group switching unit; 4. Measurement unit; 5. Discharge / collection and backflow prevention unit; 6. Controller and communication unit; 7. Cloud platform / virtual sensing area collaboration module; 8. Dual power supply unit; 9. Autonomous calibration and fluid processing unit; 11. Sealing and isolation components; 12. Shallow water intake port; 13. Water intake pipeline; 14. Filtration unit; 22. Check valve; 31. Water inlet selection valve; 32. Renewal / waste liquid switching valve; 33. Measurement branch switching valve; 34. Quality control / reagent injection valve; 35. Cleaning and maintenance valve; 36. Drain / fallback valve; 51. Waste liquid discharge pipeline; 52. Sealed collection container; 53. High-level air disconnector; 54. Liquid level alarm. Detailed Implementation

[0020] The following detailed description of a specific embodiment of the present invention is provided in conjunction with the accompanying drawings. However, it should be understood that the scope of protection of the present invention is not limited to the specific embodiment.

[0021] In the description of the present invention, it should be understood that the orientation or positional relationships indicated by the terms "center", "longitudinal", "transverse", "length", "width", "thickness", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", "axial", "radial", "circumferential" and the like are based on the orientation or positional relationships shown in the accompanying drawings, which are only for the convenience of describing the technical solutions of the present invention and simplifying the description, and do not indicate or imply that the referred device or element must have a specific orientation, be constructed and operated in a specific orientation, and thus cannot be construed as a limitation on the present invention.

[0022] In the description of the embodiments of the present invention, unless otherwise specified, the meaning of "a plurality of" refers to two or more.

[0023] Example 1 As shown in Figure 4 the present embodiment provides an adaptive pumping and drainage control method for groundwater monitoring. This method realizes adaptive optimization of the monitoring process under complex aquifer conditions by constructing a logical chain of "test - classification - HSLF switching - recovery fitting - closed loop".

[0024] In the stage of test pumping and parameter adaptive configuration, test pumping is carried out with a test flow rate for a limited time, the initial static water level and the water level response data during the pumping process are collected, the water yield capacity parameter of the aquifer is calculated, the water yield capacity of the aquifer is classified according to the water yield capacity parameter, and the corresponding pumping parameters are configured according to the classification result. The pumping parameters include rapid update flow, low-disturbance sampling flow, upper limit of drawdown rate and maximum drawdown, and the rapid update flow is greater than the low-disturbance sampling flow, and the water level drawdown rate corresponding to the low-disturbance sampling flow is not greater than the upper limit of the drawdown rate. Specifically, due to the great difference in hydrogeological conditions of different monitoring wells, the pumping strategy with fixed parameters is very likely to cause dry burning of low-permeability aquifers or insufficient update of high-permeability aquifers. In this embodiment, through short-time test pumping, the specific water yield SC is calculated as the water yield capacity parameter of the aquifer by using the measured average flow and stable drawdown, so as to quantitatively characterize the water supply performance of the aquifer. According to the SC value, the aquifer is divided into different grades such as severe low water yield, moderate low water yield and normal water yield, and then differentiated pumping strategies are matched.

[0025] When SC < SC1, it is determined as the severe low water yield grade; when SC1 ≤ SC < SC2, it is determined as the moderate low water yield grade; when SC ≥ SC2, it is determined as the non-low water yield or normal water yield grade. SC1 is preferably 0.03 to 0.08 L / min / m, and SC2 is preferably 0.15 to 0.30 L / min / m. The pumping parameters corresponding to the classification results are shown in Table 1:

[0026] Table 1: Pumping parameters corresponding to classification results As can be seen, for aquifers with severely low outflow rates, the system automatically configures a smaller, less disturbing sampling flow rate and a stricter upper limit on the drawdown rate to prevent a sudden drop in water level; while for conventional aquifers, the flow rate can be appropriately increased to shorten the monitoring cycle. This hierarchical configuration mechanism based on measured data fundamentally solves the problem that a single mode cannot adapt to varying geological environments, laying the foundation for subsequent safe and efficient operation. It should be noted that the pumping parameters corresponding to the above-mentioned results are merely exemplary. Different technicians may have different understandings of the pumping parameters and may adjust them; this invention does not limit such adjustments.

[0027] During the rapid flushing and renewal phase, a rapid renewal flow rate is used to pump out the stagnant water column in the well casing, the water retained in the shallow intake isolation section, and the residual water in the measurement front-end passage. Water level data and pump operating data are collected in real time during the pumping process. Specifically, the rapid renewal flow rate is set to a relatively high velocity, and its core function is "cleaning" rather than "sampling." In this phase, the main task of the water flow is to overcome pipeline resistance and dead volume effects, quickly replacing the stagnant water in the well that has undergone stratification, deterioration, or adsorption of contaminants due to long-term stagnation, as well as the residual water in the intake pipeline, valve assembly, and measurement unit front end. High-velocity flushing effectively eliminates the interference of historical memory effects on the monitoring results. Simultaneously, the real-time collected water level and pump operating data are not only used for safety monitoring but also provide crucial data support for subsequent judgments on switching timing. It should be understood that the specific value of the rapid renewal flow rate is not fixed but dynamically determined based on the aforementioned grading results to ensure the most efficient cleaning effect within a safe range.

[0028] During the intelligent switching and representative sampling phase, water level change trend parameters are calculated based on real-time water level data. Combined with pump operating data, it is determined whether the well casing is approaching the formation pressure equilibrium point. When approaching the equilibrium point, the pumping flow rate is smoothly switched from a rapid update flow rate to a low-disturbance sampling flow rate. Low-disturbance pumping is then performed using the configured low-disturbance sampling flow rate to obtain representative formation water from the target layer. Specifically, the "formation pressure equilibrium point" refers to a state where the water column pressure in the well and the recharge pressure of the surrounding aquifer tend to be in dynamic equilibrium. At this point, the drawdown rate slows significantly, indicating that the stagnant water in the well has been largely replaced by fresh formation water, and continuing high-flow-rate pumping no longer has a cleaning effect; instead, it may cause cavitation or the intake of water from non-target layers. Therefore, using this physical state as the trigger point for switching from high-flow-rate cleaning to low-flow-rate sampling is crucial. When the system determines that it is approaching the equilibrium point, it automatically and smoothly transitions the flow rate to the low-disturbance sampling flow rate. The low-disturbance sampling flow rate is much smaller than the rapid update flow rate, and its function focuses on "representative sampling," that is, maintaining a stable laminar flow state while minimizing disturbance to the flow field, so that the water sample entering the measurement unit can truly reflect the water quality characteristics of the target monitoring layer. This HSLF (High Stress-Low Flow) two-stage strategy cleverly balances update efficiency and sampling representativeness, avoiding the shortcomings of traditional methods that sacrifice one aspect for the other.

[0029] In the closed-loop stage of pump recovery and hydraulic parameter inversion, pumping stops when the maximum drawdown is reached, and water level rise data is continuously collected throughout the entire period. A stable recovery window is determined based on this data, and curve fitting is performed on the corresponding water level rise data. The hydraulic parameters of the aquifer with the corresponding confidence level are output based on the curve fitting results, thereby adjusting the pumping parameters for the next monitoring cycle. Specifically, the water level recovery process after pump shutdown contains rich aquifer hydraulic information. The system does not simply wait for the water level to recover, but actively identifies stable recovery windows that satisfy the quasi-steady flow assumption. It uses the Cooper-Jacob recovery method or exponential recovery model to fit the data within the window, deriving the conductivity coefficient T, permeability coefficient K, or equivalent hydraulic parameters. More importantly, this embodiment establishes a differentiated feedback adjustment mechanism based on confidence level: if the confidence level of the fitting result is high, the pumping parameters for the next cycle are significantly adjusted to approach the optimal operating condition; if the confidence level is medium, only key parameters are fine-tuned; if the confidence level is low, the original parameters are maintained and encrypted monitoring is triggered. This closed-loop mechanism of "monitoring-inversion-optimization" enables the system to continuously "learn" and adapt to local hydrogeological characteristics as it accumulates operating time, significantly improving the reliability of long-term online monitoring data and its environmental adaptability.

[0030] Example 2 Based on Example 1, this example further specifies the formation pressure balance point identification logic in the intelligent switching determination process. Water level change trend parameters include the derivative of the drawdown curve and / or the drawdown per unit pumping volume increment. The method for determining proximity to the formation pressure balance point includes: when the derivative of the drawdown curve is lower than a preset derivative threshold and the duration is not less than a preset duration, and / or the drawdown per unit pumping volume increment is lower than a preset volume threshold, and simultaneously, when the pump operating data meets stability constraints, it is determined that the formation pressure balance point is approaching.

[0031] Specifically, to accurately capture the moment when the water column pressure in the well and the aquifer recharge pressure tend to reach dynamic equilibrium, this invention introduces a multi-dimensional combination of criteria. Among these, the derivative of the drawdown curve, |dH / dt|, reflects the rate of change of water level over time and is a direct indicator of flow stability. However, the groundwater monitoring environment is complex, and the water level signals collected by sensors are often accompanied by noise or short-term pulse interference. Relying solely on the instantaneous derivative can easily lead to misjudgments. Therefore, the derivative of the drawdown curve is obtained through sliding window regression or smoothed difference. For example, a sliding window with a width of 30s to 180s is used to perform linear regression fitting on continuous water level data, and the slope of the fitted line is used as the derivative value at the current moment; or a Savitzky-Golay filter is used for smoothed difference processing. This processing method can effectively filter out high-frequency random noise, retain the true trend of water level changes, and significantly improve the robustness of the judgment. The preset derivative threshold α is preferably 0.5cm / min to 2.0cm / min, and the preset duration T1 is preferably 30s to 300s. The specific values ​​of α and T1 can be adaptively adjusted according to the aquifer type and sensor accuracy: for aquifers with poor permeability and slow water level response, a smaller α value and a longer T1 should be selected to avoid premature switching; while for high-sensitivity sensors or high-permeability aquifers, the threshold can be appropriately relaxed to improve the response speed.

[0032] As an auxiliary criterion, the drawdown per unit pumped volume increment ΔH / ΔV is introduced to compensate for the shortcomings of a purely time-domain criterion. This parameter is defined as the ratio of the drawdown increment to the corresponding cumulative pumped volume increment over a certain period. Its physical meaning lies in eliminating the interference of the time factor and directly characterizing the aquifer's unit volume replenishment capacity under the current pumping intensity. When ΔH / ΔV is lower than the preset volume threshold β, it means that the water level drop caused by each unit volume of water pumped is negligible, indicating that the aquifer replenishment is close to the pumping volume, and the system has essentially entered a quasi-steady state. β is preferably between 0.01 m / L and 0.20 m / L. Combining ΔH / ΔV with |dH / dt| constitutes a "time-volume" dual verification mechanism, effectively preventing the failure of a single criterion due to pump flow fluctuations or sensor drift.

[0033] More importantly, to prevent "false balance" caused by equipment failure, this embodiment sets strict stability constraints on pump operation. These constraints include ensuring the flow fluctuation coefficient does not exceed a threshold and that the correlation between pump current and flow rate conforms to the normal operating range. In actual operation, if abnormalities such as air cavitation, filter blockage, or impeller wear occur, the pump may experience a sudden drop or severe fluctuation in flow rate. At this time, the water level drawdown rate will also slow down or even stop, which can easily be misjudged as reaching the formation pressure equilibrium point. Therefore, the system calculates the flow fluctuation coefficient CV_Q (i.e., the ratio of the standard deviation of flow rate to the average value) in real time and requires it not to exceed a preset threshold (e.g., 0.10 to 0.30). Simultaneously, it monitors the correlation between pump current I(t) and flow rate Q(t). If either deviates from the calibrated normal operating curve (e.g., abnormally high current without increased flow rate, or violent oscillations in sync with the flow rate), the operating condition is determined to be unstable, the equilibrium point determination logic is immediately locked, and safety protection is triggered. This design ensures that traffic switching is only performed when the device is in good working order, fundamentally avoiding the risk of data distortion or device damage caused by erroneous switching.

[0034] In summary, this embodiment employs a triple verification mechanism of "derivative trend + volumetric efficiency + operational health" to precisely define the trigger boundary for a smooth transition from rapidly updated flow rate to low-disturbance sampling flow rate. This not only ensures that the stagnant water in the well is fully replaced before entering the low-disturbance sampling stage, but also ensures that the switching process remains within a safe and controllable operating range, providing solid algorithmic support for obtaining high-quality, representative formation water samples.

[0035] Example 3 Building upon the pump shutdown recovery and hydraulic parameter estimation steps in Example 1, this example further specifies the accuracy control and adaptive closed-loop adjustment mechanism for hydraulic parameter inversion. The method for determining the stable recovery window based on water level rise data includes: identifying the time interval satisfying the quasi-steady flow assumption of well flow theory from the full-time water level rise data according to preset recovery criteria, as the stable recovery window. Specifically, the water level recovery process after pump shutdown is not applicable to the entire parameter estimation using standard well flow formulas. In the initial recovery phase, the water level rise rate is often extremely rapid and does not conform to the formation seepage law due to the wellbore reservoir effect; while in the later recovery phase, if the time is too long, it may be affected by boundary conditions or regional background water levels. Therefore, accurately identifying the "stable recovery window" is a prerequisite for ensuring the accuracy of parameter inversion. The recovery criteria include at least one of the following: the error in restoring the water level to the reference water level does not exceed 0.02m to 0.10m; the absolute value of the water level change rate does not exceed 0.1cm / min to 0.5cm / min and the duration is not less than 120s to 600s; the real-time fitting residual of the water level recovery curve is less than a preset threshold and remains stable across multiple consecutive sampling points. In practical applications, these three criteria are usually used in combination to construct a multi-verification mechanism. For example, firstly, the water level change rate criterion is used to screen out candidate intervals with gradual changes, excluding segments with drastic fluctuations; then, the fitting residual is checked within this interval to ensure that the data points fall within the allowable deviation band of the theoretical curve, eliminating local outliers; finally, the reference water level error criterion is used to confirm whether the degree of recovery has reached an acceptable range. The reference water level... It can be the initial static water level of the current cycle, the background water level of the recent effective cycle, or the reference water level after regional trend correction; allowable error The preferred range is 0.02m to 0.10m, which balances sensor resolution with practical engineering requirements; water level change rate threshold. Preferably, the speed is 0.1 cm / min to 0.5 cm / min, and the duration is... The optimal timeframe is between 120s and 600s. This combined identification strategy can effectively filter out non-stationary data, providing a high-quality data foundation for subsequent fitting.

[0036] After determining the stable recovery window, curve fitting methods include: using at least one selected from the unit gradient equation, the Cooper-Jacob recovery method, or the exponential recovery model to fit the water level recovery data within the stable recovery window. Specifically, different hydrogeological conditions correspond to different well flow theory models. For homogeneous, isotropic, and infinitely extended confined aquifers, the Cooper-Jacob recovery method is a classic and robust choice, which involves fitting the residual drawdown... Compared to time To solve for the transmission coefficient, a linear regression is performed on the logarithmic relationship. For aquifers with overflow recharge or constant head boundaries, the exponential recovery model better describes the asymptotic recovery characteristics of water level; however, in emergency scenarios involving rapid on-site assessment or low data quality, the unit gradient equation... It can be used as a rough estimation method for permeability coefficient. The system can automatically select the best-fitting model based on historical goodness of fit, aquifer type label, or recovery curve morphology, or select the best model after parallel calculation of multiple models, thereby avoiding systematic bias caused by model misuse.

[0037] To ensure the reliability of the inversion results, curve fitting also includes validity verification. Validity verification criteria include: the regression coefficient of determination of the curve is not less than a preset value, and the root mean square error is not greater than a preset value; the water level recovery curve within the fitting interval is monotonically increasing without abnormal fluctuations; and there are no cavitation, blockages, or abnormal valve positions during the pumping process. Specifically, statistical indicators alone are insufficient to determine the physical authenticity of hydraulic parameters. (Regression coefficient of determination) The root mean square error (RMSE) (preferably not less than 0.90 to 0.98) and the physical mean square error (RMSE) (preferably not greater than 0.02 m to 0.05 m) reflect the degree of agreement between the data and the model; the monotonically increasing constraint is based on the physical law of groundwater recovery, and any non-monotonic "bounce" or "drop" suggests measurement interference or equipment failure; while the operational normality check correlates hydraulic analysis with electromechanical status, eliminating the risk of water level distortion caused by cavitation, blockage, or abnormal valve position being misinterpreted as formation characteristics. Only data that pass the triple verification of statistical tests, physical constraints, and equipment health checks are considered valid samples.

[0038] Based on the above verification results, the hydraulic parameters of the aquifer corresponding to the confidence level are output according to the curve fitting results, including: when the fitting results meet all validity verification criteria, the high-confidence conductivity and permeability coefficients are output; when the fitting results partially meet the validity verification criteria, the medium-confidence conductivity and permeability coefficients are output; when the fitting results do not meet the validity verification criteria, the low-confidence equivalent hydraulic parameters are output. Furthermore, the pumping parameters for the next monitoring cycle are adjusted according to the corresponding confidence level of the aquifer hydraulic parameters, including: high-confidence hydraulic parameters can be fully adjusted, with an adjustment range not exceeding ±30%; medium-confidence hydraulic parameters only require minor adjustments to the low-disturbance sampling flow rate and pump shutdown recovery time, with an adjustment range not exceeding ±15%; low-confidence hydraulic parameters do not require any adjustment to the pumping parameters, maintaining the effective setpoint of the previous cycle unchanged, and triggering intensified monitoring. This tiered adaptive adjustment strategy is the core defense point that distinguishes this invention from the existing "one-size-fits-all" parameter update mode. Specifically, high confidence indicates that the current data quality is extremely high and the model fit is good. The system should make full use of this information to quickly approach the optimal operating condition. Therefore, a relatively large adjustment (±30%) of all parameters is allowed to improve convergence efficiency. Medium confidence indicates that the data has some flaws or uncertainties. In this case, large adjustments are likely to cause system oscillations. Therefore, adjustments are only made to key sensitive parameters (such as low-disturbance sampling flow). The system conservatively fine-tunes the parameters (including recovery time) by ±15%, striking a balance between exploration and safety. Low confidence levels warn that the current inversion results are unreliable; forcing their application would only worsen the monitoring effect. Therefore, the system locks the parameters unchanged and triggers encrypted monitoring to accumulate more data for subsequent diagnosis. This differentiated feedback mechanism based on evidence strength ensures the system's rapid self-learning capability under ideal conditions while also providing robustness and stability in complex and harsh environments, effectively mitigating the risk of the entire monitoring strategy failing due to a single erroneous inversion.

[0039] Example 4 Based on the adaptive pumping control process described in Examples 1 to 3, this example further introduces a safety constraint mechanism that runs through the entire pumping process to ensure the system's operational reliability and equipment safety in long-term unattended environments. This safety constraint includes: real-time monitoring of water level, flow rate, and pump current throughout the pumping process; immediately executing speed reduction, pump stoppage, or alarm operations when the maximum pumping duration, minimum allowable water level, or maximum pump current threshold is triggered; and executing venting or short-term pump stoppage protection actions when cavitation, idling, or blockage risks are detected.

[0040] Specifically, safety constraints constitute a low-level protection system independent of the normal adaptive control logic, and their execution priority is higher than the aforementioned business logic such as probing grading, traffic switching, and parameter inversion. This means that regardless of the monitoring stage the system is in, once a safety boundary is reached, the controller will unconditionally suspend the current business task and prioritize responding to safety commands. This design is an important feature that distinguishes engineered online monitoring equipment from laboratory analytical instruments, aiming to cope with unforeseen anomalies that may occur under complex field conditions.

[0041] Regarding hard threshold protection, the system sets absolute safety boundaries for key state variables. For example, the maximum pumping duration threshold is used to prevent the pump from overheating due to prolonged dry running caused by unexpected attenuation of aquifer recharge capacity or pipeline leakage; the minimum allowable water level threshold is usually set at a certain safe distance above the shallow intake or the top of the filter pipe to prevent air intake and cavitation caused by excessively low water levels, or abnormal increase in turbidity caused by disturbance of sediment at the bottom of the well; the maximum pump current threshold is directly related to the motor load status and is used to prevent electrical faults caused by mechanical jamming, abnormal voltage, or winding short circuits. When any of the above hard indicators exceeds the limit, the controller does not perform complex trend analysis or delayed confirmation, but directly cuts off the pump source or forces a speed reduction with a millisecond-level response speed, and simultaneously generates an alarm record. This instantaneous circuit breaker mechanism can effectively prevent irreversible physical damage to the equipment.

[0042] In terms of identifying soft risks, the system uses multi-source data fusion analysis to identify potential risks that have not yet reached the hard threshold but have already shown signs of failure. For example, the determination of cavitation risk does not rely solely on a single pressure sensor reading, but comprehensively monitors the pump inlet pressure pulsation frequency, flow fluctuation coefficient CV_Q, and the high-frequency noise component of the pump current; when these three exhibit a specific correlated abnormal pattern, even if the water level is still within the safe range, the system determines that there is a potential cavitation risk. For idling risk, it is mainly identified by comparing whether the pump current is lower than the no-load reference value and whether the flow meter reading remains zero. For blockage risk, it is determined by analyzing the matching relationship between the increasing trend of the pressure difference before and after the filter unit and the flow rate attenuation ratio. For different soft risks, the system executes differentiated flexible protection actions: when cavitation or bubble interference is detected, the system automatically triggers the venting and emptying procedure, releasing accumulated gas by briefly opening the venting valve; when slight blockage or overheating tendency is detected, a short-term pump stop protection is executed, giving the equipment a time window for cooling or stress relief, and attempting to restart after conditions are restored. This feature extraction-based proactive defense strategy is better suited to the gradual faults commonly found in groundwater monitoring compared to simple threshold alarms.

[0043] Furthermore, the security constraint mechanism is closely linked to the system's anomaly diagnosis system. For example... Figure 8In the abnormal scenario matrix shown, the triggering conditions for Class E anomalies (valve position / backflow risk) and Class C anomalies (blockage / scaling / biofilm) often originate directly from the monitoring data of the safety constraint module. When a safety constraint is triggered, the system not only executes immediate hardware protection actions but also packages and marks relevant process parameters (such as water level curves, current waveforms, and differential pressure change rates at the trigger time) as abnormal evidence for subsequent diagnostic engine classification. For example, if blockage risk protection is triggered due to flow attenuation, the system will automatically associate with the Class C anomaly handling procedure and attempt to perform backwashing maintenance after a safe shutdown; if backflow risk protection is triggered due to inconsistent valve position feedback, it will associate with the Class E anomaly handling procedure, forcibly locking the external discharge path and entering a safe standby state. This integrated design of "safety protection - diagnostic tracing - automatic handling" ensures that the system can not only "stop" when encountering anomalies but also "record" and attempt to "repair" them, thereby maximizing the data continuity and equipment availability of long-term online monitoring.

[0044] Example 5 like Figure 1 As shown, this embodiment provides a groundwater monitoring pumping system, which provides a specific hardware carrier and actuator for the adaptive pumping control method described in Embodiments 1 to 4. The pumping system includes a pumping update unit 2, a water level measurement unit, a process sensing unit, a valve group switching unit 3, a measurement unit 4, a discharge / collection and anti-backflow unit 5, and a controller 6. In terms of physical connection and communication topology, the controller 6 is communicatively connected to the pumping update unit 2, the water level measurement unit, the process sensing unit, the valve group switching unit 3, and the measurement unit 4, forming the system's central nervous system. In terms of water circuit connection, the pumping update unit 2 is connected to the valve group switching unit 3, and the valve group switching unit 3 is connected to the measurement unit 4 and the external discharge pipeline, forming a controllable fluid distribution network.

[0045] Specifically, the pumping and updating unit 2 receives instructions from the controller 6 and performs trial pumping, rapid update flow pumping, smooth flow switching, and low-disturbance sampling flow low-disturbance pumping actions, stopping operation when the maximum drawdown depth is reached. For example... Figure 2As shown, to achieve precise capture of shallow groundwater, the system also includes a downhole shallow water intake and isolation unit 1, which includes a sealing isolation component 11, a shallow water intake 12, a water intake pipeline 13, and a filtration unit 14, all installed within the monitoring well casing. The sealing isolation component 11 can be a packer or an expansion sealing ring arranged vertically at intervals, creating a relatively independent shallow water intake isolation section within the well casing, effectively preventing the vertical mixing of water from non-target layers. The shallow water intake 12 is located within this isolation section and connects to the surface pumping and renewal unit 2 via the water intake pipeline 13, which passes through the sealing isolation component 11. The filtration unit 14 is installed outside the water intake or at the pipeline inlet to intercept sand and suspended solids, protecting delicate downstream components. This structural design ensures that the water sample entering the system truly originates from the target monitoring layer, laying a physical foundation for the accuracy of subsequent water quality and hydraulic parameters.

[0046] Valve group switching unit 3 is used to coordinate water circuit on / off switching to ensure water delivery and pipeline water renewal. For example... Figure 3 As shown, this unit achieves automated management of complex flow paths through a combination of multiple valves. Specifically, it includes an inlet water selection valve 31, a refresh / waste liquid switching valve 32, a measurement branch switching valve 33, a quality control / reagent injection valve 34, a cleaning and maintenance valve 35, a drain / backflush valve 36, and a check valve or anti-crossflow valve 37. The inlet water selection valve 31 is located between the intake pipeline 13 and the main flow path at the station end, and can be used as a well selection manifold valve in a multi-well shared system. The refresh / waste liquid switching valve 32 is located at the outlet of the pumping and refreshing unit 2, responsible for directing the initial flushing water into the waste liquid branch to avoid contaminating the measurement path. The measurement branch switching valve 33 controls the timing of representative formation water entering the measurement unit 4. The quality control / reagent injection valve 34 connects to blank solution, standard solution, or colorimetric reagent storage tanks, supporting online quality control and chemical analysis. The cleaning and maintenance valve 35 connects to the cleaning fluid or backflushing branch to address blockages or scaling. The drain / backflush valve 36 is located at the lowest point of the flow path or at the outlet of the measurement unit 4, used to drain residual liquid. Check valves or anti-crossflow valves 37 are distributed at key nodes in each branch to prevent cross-contamination between different fluids. The controller 6 coordinates the opening and closing sequence of the above valves to precisely execute the state machine flow of "flushing-sampling-measurement-quality control-draining", ensuring the airtightness and sequence of the entire process.

[0047] The emission / collection and backflow prevention unit 5 is a key line of defense ensuring the long-term unattended safety of the system. For example... Figure 3As shown, the unit includes a waste liquid discharge pipeline 51, a closed collection container 52, a high-level air disconnector 53, a liquid level alarm 54, and a check valve 22. The high-level air disconnector 53 is located at the highest point of the waste liquid discharge pipeline 51, with its installation elevation strictly higher than the outlet of the measuring unit 4 and the highest liquid level of the closed collection container 52. When the system is in standby mode or when negative pressure is generated at the discharge end, the high-level air disconnector 53 can introduce air to disrupt the continuous liquid column, fundamentally cutting off the physical conditions for siphon formation. Simultaneously, the check valve 22 only allows unidirectional discharge of waste liquid, automatically closing under reverse pressure differential. The two work together to form a "dual gas-liquid blocking" mechanism, completely eliminating the risk of external liquid or waste liquid flowing back into the measuring passage or even the monitoring well. The liquid level alarm 54 monitors the liquid level status of the closed collection container 52 in real time to prevent overflow accidents.

[0048] The water level measurement unit is used to collect initial static water level, water level response data and water level data during pumping, as well as water level recovery data throughout the entire time period after pump shutdown, and transmits the data to controller 6; the process sensing unit is used to collect pump operating data and upload it to controller 6. These sensors constitute the system's sensing layer, providing real-time feedback for adaptive control. Measurement unit 4 is used to receive representative formation water from the target stratum. It can integrate multiple sensors such as pH, conductivity, turbidity, and dissolved oxygen, and preferably adopts a near-zero dead volume flow path design to reduce retention and cross-contamination.

[0049] Controller 6, as the system's decision-making core, calculates aquifer discharge capacity parameters based on collected data, classifies aquifer discharge capacity, and adaptively configures rapidly updated flow rate, low-disturbance sampling flow rate, upper limit of drawdown rate, and maximum drawdown. It calculates water level change trend parameters based on water level data, determines the formation pressure balance point based on pump operating data, and controls flow switching. It also determines the stable recovery window based on water level rise data, performs curve fitting on the corresponding data, outputs aquifer hydraulic parameters with corresponding confidence levels, and adjusts the pumping parameters for the next monitoring cycle accordingly. In its implementation, controller 6 incorporates an embedded computing module and storage unit, running the adaptive algorithm program described in Examples 1 to 4. It can not only process high-frequency sampled water level and operating condition signals but also dynamically update the control strategy based on the inversion results, achieving a leap from "passive execution" to "active optimization." Furthermore, the system can be equipped with a dual-energy power supply unit 8 to support operation in remote areas, an autonomous calibration and fluid processing unit 9 for maintenance-free quality control, and a cloud platform / virtual sensing regional collaboration module 7 for big data analysis and regional early warning. Through the organic integration of the above hardware units, this embodiment constructs a complete, reliable, and self-evolving online groundwater monitoring system, which fully supports the implementation of the technical solutions in the aforementioned method claims.

[0050] Example 6 Based on the aforementioned embodiments, this embodiment further demonstrates the comprehensive application effect of the groundwater monitoring adaptive pumping control method and system under various typical hydrogeological conditions and complex operation and maintenance scenarios. The aim is to verify the robustness of the technical solution in actual engineering environments, its ability to judge data validity, and its traceability throughout the entire process.

[0051] In monitoring scenarios targeting aquifers with low water yield, the system demonstrated excellent adaptive adjustment capabilities. Taking a monitoring well as an example, the well diameter D is 110 mm, and the initial static water level is approximately 10 m deep. During the trial pumping phase, the controller set the trial flow rate Q_test to 0.30 L / min and the time limit t_test to 120 s. The measured stable drawdown ΔH_test was 3.0 m. The specific water yield SC was calculated as 0.10 L / min / m using the formula SC = Q_test / ΔH_test. Based on the preset classification threshold, this value falls within the medium-low water yield range. Based on this classification result, the system automatically configured a rapid update flow rate Q1 of 1.2 L / min for rapid flushing and a low-disturbance sampling flow rate Q2 of 0.15 L / min for low-disturbance sampling. Simultaneously, the upper limit of the drawdown rate α2 was set to 0.5 cm / min, and the maximum drawdown ΔH_max was set to 2.0 m to prevent excessive pumping leading to dry burning or cavitation. During the subsequent rapid flow rate update and pumping process, the system calculates the derivative of the drawdown curve, |dH / dt|, and the drawdown per unit pumping volume increment, ΔH / ΔV, in real time. When |dH / dt| calculated using a 60-s sliding window regression is less than 1.0 cm / min and lasts for 120 seconds, and ΔH / ΔV is less than 0.05 m / L, the controller determines that the well is close to the formation pressure equilibrium point and smoothly switches to a low-disturbance sampling flow rate Q2 for representative sampling. During the pump shutdown and recovery phase, the system identifies a stable recovery window that satisfies the quasi-steady flow assumption and uses the Cooper-Jacob recovery method to linearly fit the relationship between the residual drawdown and the logarithm of time. When the regression determination coefficient R² ≥ 0.95 and the root mean square error RMSE ≤ 0.02 m, the system outputs high-confidence conductivity coefficient T and permeability coefficient K, and accordingly controls the adjustment range of pumping parameters within ±30% in the next monitoring cycle. This complete numerical transfer process demonstrates that the present invention can dynamically optimize the monitoring strategy based on the actual water supply performance of the aquifer, effectively solving the technical problem of balancing renewal efficiency and sampling representativeness under low permeability conditions.

[0052] In contrast, in a conventional outflow aquifer scenario, when the calculated specific outflow SC ≥ 0.20 L / min / m, the system automatically adjusts the control strategy to adapt to the faster recharge capacity. For example, the rapid update flow rate Q1 is adjusted to 0.8 L / min, the low-disturbance sampling flow rate Q2 is adjusted to 0.2 L / min, the upper limit of the drawdown rate α2 is relaxed to 1.0 cm / min, and the target drawdown ΔH* is set to 1.0 m. This differentiated parameter configuration indicates that the present invention does not employ a single fixed control mode, but rather establishes a wide-range adaptive mechanism covering aquifers from extremely low permeability to high permeability, ensuring the timeliness and safety of monitoring under different hydrogeological conditions.

[0053] In the identification and early warning of suspected pollution incidents, this invention introduces a regional collaborative monitoring mechanism to suppress false alarms. For example... Figure 6 As shown, the cloud platform integrates online monitoring data from multiple sites with external data such as meteorological and remote sensing data, and uses a virtual sensing model to invert the spatial dynamics of shallow groundwater levels and the background field distribution in the region. When a single station detects an anomaly in water quality factors, the system calculates the regional consistency index (CI) to assist in decision-making. For example, if the conductivity of a station increases by more than 10% of the threshold and the turbidity does not increase significantly at the same time, and the calculated CI is ≥0.7, it indicates that the anomaly is highly consistent with the regional background trend. The system classifies it as a "high-confidence pollution anomaly," automatically upgrades the alarm level, and triggers encrypted monitoring. Conversely, if the CI is <0.3 or accompanied by quality control failure, it is preferentially judged as instrument drift or abnormal operating conditions, triggering retesting and maintenance procedures instead of pollution alarms. This discrimination logic based on spatiotemporal correlation analysis effectively distinguishes between real environmental pollution signals and the systematic deviations of the equipment itself, significantly improving the reliability of online monitoring data and the efficiency of regulatory response.

[0054] To ensure the legal validity of data from long-term unattended operation and the traceability of maintenance, the system has established a diagnostic evidence chain based on an anomaly scenario matrix. For example... Figure 8As shown, the controller incorporates seven anomaly classification rules (AG), solidifying the "trigger condition - handling action - quality mark" into an immutable logical association. For example, when a discrepancy is detected between the valve position feedback and the target state, or when an abnormal rise in the discharge level occurs during standby, a Class E (valve position / backflow risk) anomaly is triggered. The system immediately cuts off the discharge path, switches to closed collection, and enters safe standby, while simultaneously marking the data for this cycle as "invalid." When the differential pressure in the measurement path increases and the flow rate decreases beyond the threshold, a Class C (blockage / scaling) anomaly is triggered, automatically performing backflushing or chemical cleaning maintenance. Each monitoring cycle's diagnostic report includes anomaly classification codes, original trigger factor values, threshold comparison results, handling action records, and associated process curve data packets, all stored using hash verification or digital signatures. This structured evidence chain design ensures that any data anomaly or equipment failure can be traced back to a specific physical cause and handling point, meeting the stringent environmental regulatory requirements for data integrity and audit trails.

[0055] Furthermore, to address the monitoring needs of emerging trace pollutants such as PFAS, the system also integrates a material blank verification process. After the equipment is put into operation or the flow path components are replaced, the controller automatically calls upon verified blank water to circulate within the measurement path, collecting baseline values ​​of PFAS-related indicators or alternative indicators, and binding and storing the verification results with the material configuration file. If the blank value deviates from the preset threshold, the system will prompt the replacement of the flow path material or the execution of a more stringent cleaning procedure, thereby reducing background interference from materials such as fluoropolymers at the source. Combining the aforementioned adaptive drainage, regional collaborative diagnosis, and evidence storage mechanisms, this embodiment constructs a comprehensive, multi-layered technical verification system, fully demonstrating the high practical value and promising prospects of this invention in complex and ever-changing groundwater shallow pollution monitoring scenarios.

[0056] It should be noted that any parts not disclosed or specifically described in this invention are existing technology or conventional configurations, and their specific structures and working principles will not be elaborated further. In this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0057] Although embodiments of the present invention have been disclosed above, they are not limited to the applications listed in the specification and embodiments. It can be applied to various fields suitable for the present invention. Other modifications can be readily implemented by those skilled in the art. Therefore, without departing from the general concept defined by the claims and their equivalents, the present invention is not limited to the specific details and examples shown and described herein.

Claims

1. An adaptive pumping control method for groundwater monitoring, characterized in that, include: A trial pumping operation is conducted for a limited time using a trial flow rate. Initial static water level and water level response data during the pumping process are collected. Aquifer discharge capacity parameters are calculated. The aquifer discharge capacity is classified according to the aquifer discharge capacity parameters. Corresponding pumping parameters are configured based on the classification results. The pumping parameters include rapid update flow rate, low disturbance sampling flow rate, upper limit of drawdown rate, and maximum drawdown. The rapid update flow rate is greater than the low disturbance sampling flow rate, and the drawdown rate corresponding to the low disturbance sampling flow rate is not greater than the upper limit of drawdown rate. Pumping is carried out by rapidly updating the flow rate, updating the static water column in the well pipe, the water retained in the shallow water intake isolation section and the residual water in the measurement front-end passage, and collecting water level data and pump operating data in real time during the pumping process. Calculate water level change trend parameters based on real-time water level data, and determine whether the well casing is close to the formation pressure balance point by combining pump operating data. When it is close to the formation pressure balance point, smoothly switch the pumping flow rate from rapid update flow rate to low disturbance sampling flow rate, and perform low disturbance pumping with the configured low disturbance sampling flow rate to obtain representative formation water of the target layer. Pumping is stopped when the maximum drawdown is reached. Water level rise data is continuously collected throughout the entire period. A stable recovery window is determined based on the water level rise data. Curve fitting is performed on the water level rise data corresponding to the stable recovery window. The hydraulic parameters of the aquifer with the corresponding confidence level are output based on the curve fitting results, thereby adjusting the pumping parameters for the next monitoring cycle.

2. The adaptive pumping control method for groundwater monitoring according to claim 1, characterized in that, The water level change trend parameters include the derivative of the water level drawdown curve and / or the drawdown per unit pumping volume increment; the method for determining the near formation pressure equilibrium point includes: When the derivative of the drawdown curve is lower than a preset derivative threshold and the duration is not less than a preset duration, and / or the drawdown per unit pumping volume increment is lower than a preset volume threshold, and the pump operating data meets the stability constraints, it is determined that the formation pressure equilibrium point is approaching.

3. The adaptive pumping control method for groundwater monitoring according to claim 2, characterized in that, The derivative of the drawdown curve is obtained through sliding window regression or smoothing difference. The stability constraints include that the flow fluctuation coefficient does not exceed the threshold and that the correlation characteristics between pump current and flow rate are within the normal operating range.

4. The adaptive pumping control method for groundwater monitoring according to claim 1, characterized in that, The method for determining the stable recovery window based on water level rise data includes: Based on a preset recovery criterion, time intervals that satisfy the quasi-steady flow assumption of well flow theory are identified from the full-time water level rise data as stable recovery windows; the recovery criterion includes at least one of the following: The error in restoring the water level to the reference level shall not exceed 0.02m to 0.10m; The absolute value of the rate of change of water level shall not exceed 0.1 cm / min to 0.5 cm / min and the duration shall not be less than 120 s to 600 s; The real-time fitting residual of the water level recovery curve is less than the preset threshold and remains stable at multiple consecutive sampling points.

5. The adaptive pumping control method for groundwater monitoring according to claim 1, characterized in that, The curve fitting method includes: using at least one selected from the unit gradient equation, the Cooper-Jacob recovery method, or the exponential recovery model to perform curve fitting on the water level recovery data within the stable recovery window.

6. The adaptive pumping control method for groundwater monitoring according to claim 1, characterized in that, The curve fitting also includes validity verification, and the validity verification criteria include: The regression coefficient of the curve is not less than the preset value, and the root mean square error is not greater than the preset value. The water level recovery curve within the fitted interval shows a monotonically increasing trend with no abnormal fluctuations. There was no cavitation, blockage, or abnormal valve position during the extraction process.

7. The adaptive pumping control method for groundwater monitoring according to claim 6, characterized in that, The step of outputting aquifer hydraulic parameters with corresponding confidence levels based on curve fitting results includes: When the fitting results meet all validity verification criteria, output the transmission coefficient and penetration coefficient with high confidence. When the fitting results partially meet the validity verification criteria, the transmission coefficient and penetration coefficient with confidence level are output. When the fitting results do not meet the validity verification criteria, output the equivalent hydraulic parameters with low confidence.

8. The adaptive pumping control method for groundwater monitoring according to claim 7, characterized in that, The adjustment of the pumping parameters for the next monitoring cycle based on the aquifer hydraulic parameters with corresponding confidence levels includes: High-confidence hydraulic parameters allow for full adjustment of pumping parameters, with an adjustment range not exceeding ±30%. For medium-confidence hydraulic parameters, only minor adjustments are made to the low-disturbance sampling flow rate and pump shutdown recovery time, with the adjustment range not exceeding ±15%. Low-confidence hydraulic parameters do not adjust any pumping parameters, maintain the valid setpoints from the previous cycle, and trigger encrypted monitoring.

9. The adaptive pumping control method for groundwater monitoring according to claim 1, characterized in that, It also includes security constraints, which include: The pumping process monitors water level, flow rate and pump current in real time. When the maximum pumping time, minimum allowable water level or maximum pump current threshold is triggered, the pumping speed is reduced, the pump is stopped or an alarm is triggered immediately. When cavitation, idling, or blockage risks are detected, exhaust or short-term pump stop protection actions are executed.

10. A groundwater monitoring pumping system, characterized in that, It includes a pumping and renewal unit, a water level measurement unit, a process sensing unit, a valve group switching unit, a measurement unit, a discharge / collection and anti-backflow unit, and a controller; The controller is communicatively connected to the pumping and updating unit, the water level measurement unit, the process sensing unit, the valve group switching unit, and the measurement unit. The water circuit of the pumping and updating unit is connected to the valve group switching unit, and the valve group switching unit is connected to the measurement unit and the external discharge pipeline. The pumping and updating unit is used to receive controller instructions and complete trial pumping, rapid update flow pumping, smooth flow switching, and low-disturbance sampling flow low-disturbance pumping actions, and stops operation when the maximum drawdown depth is reached. The water level measurement unit is used to collect initial static water level, water level response data and water level data during the pumping process, and water level recovery data throughout the entire time period after the pump stops, and transmit the data to the controller; the process sensing unit is used to collect pump operating condition data and upload it to the controller. The valve group switching unit is used to coordinate the switching of water circuit on / off to ensure water delivery and pipeline water renewal; The measurement unit is used to collect representative formation water from the target stratum. The controller is used to calculate the aquifer discharge capacity parameters based on the collected data, classify the aquifer discharge capacity, and adaptively configure the fast-updating flow rate, low-disturbance sampling flow rate, upper limit of drawdown rate, and maximum drawdown. It calculates the water level change trend parameters based on the water level data, determines the formation pressure balance point based on the pump operating data, and controls the flow rate switching. It determines the stable recovery window based on the water level rise data, performs curve fitting on the corresponding data, outputs the aquifer hydraulic parameters with corresponding confidence levels, and adjusts the pumping parameters for the next monitoring cycle accordingly.

Images