Underground water extracting and sampling device

Abstract

The invention discloses an underground water extracting and sampling device. According to the invention, through the automatic control unit and the modular design, the sampling precision and the operation efficiency are improved, and personal errors are reduced. The device can automatically complete timed and quantitative extraction, the real-time monitoring capability on the underground water flow and the water quality is enhanced, and the sampling data is ensured to be accurate and reliable. Meanwhile, the service life of the equipment in a complex environment is prolonged due to the design of a corrosion-resistant material and a low-power-consumption power supply, a water quality parameter detection function is conveniently expanded due to a modular framework, diversified requirements under different scenes are met, the maintenance cost and the operation difficulty are reduced, and by integrating a water quality self-adaptive module, the service life of the equipment is prolonged. And the intelligent level and the data reliability of the sampling process are obviously improved. The water quality self-adaption module can monitor key parameters such as turbidity and conductivity in underground water in real time, it is ensured that the collected water sample meets the monitoring standard, meanwhile, the frequency of manual repeated debugging and field maintenance is reduced, and the operation efficiency is improved.

Description

Technical Field

[0001] The invention belongs to the technical field of groundwater sampling, and in particular relates to a groundwater extraction and sampling device. Background Art

[0002] Groundwater sampling is a crucial component of environmental monitoring and hydrogeological research. Traditional groundwater sampling methods rely heavily on manual labor, resulting in low sampling efficiency, complex operation, and susceptibility to human influence. With the advancement of automation technology, some groundwater sampling devices have begun to incorporate automation features, but issues such as low sampling accuracy and an inability to achieve timed and quantitative sampling persist. Furthermore, existing automated sampling devices often rely on complex control systems, resulting in high costs and difficulty in promoting their use in small and medium-sized projects. Currently, common groundwater sampling methods include manual sampling pumps, electric sampling pumps, and sensor-based automated sampling devices. Manual sampling pumps are simple to operate but offer low sampling accuracy and efficiency. While electric sampling pumps improve sampling efficiency, they still require manual intervention and cannot be fully automated. Sensor-based automated sampling devices can achieve partial automation, but their complex structure and high cost make them difficult to meet sampling requirements in diverse scenarios.

[0003] However, the main drawbacks of existing technologies are low sampling accuracy, inability to achieve timed and quantitative sampling, complex structure, high cost, and limited scope of application. These problems restrict the further development and application of groundwater sampling technology. Summary of the Invention

[0004] The purpose of the present invention is to provide a groundwater extraction and sampling device in order to solve the above-mentioned problems.

[0005] The technical solution adopted by the present invention is as follows: a groundwater extraction sampling device, the sampling device comprising a sampling pump, a pipeline, a flow sensor, a water storage container, a timer and a control unit; The water outlet of the sampling pump is connected to the groundwater well through the inlet end of a stainless steel corrugated pipe, and the outlet end of the pipe is connected to the water inlet of the water storage container; The flow sensor is embedded in the middle section of the pipeline, and the electrode signal output end of the flow sensor is connected to the input end of the data acquisition module of the control unit through a shielded cable; The liquid level float switch signal line of the water storage container is connected to the input port of the logic control module of the control unit. The clock signal output terminal of the timer communicates with the microcontroller of the control unit via the I2C bus; The PWM signal output terminal of the execution drive module of the control unit is connected to the motor drive circuit of the sampling pump; The control unit is internally provided with a data acquisition module, a control logic module, an execution drive module, a communication interface module, a power management module and a water quality adaptive module; The water quality self-adaptation module consists of a water quality data acquisition submodule, an abnormality analysis and decision submodule, and a feedback execution submodule.

[0006] In a preferred embodiment, the sensor signal output end of the data acquisition module is connected to the SPI communication input end of the control logic module via a shielded cable, wherein the electrode signal of the flow sensor and the clock signal of the timer are respectively connected to the analog and digital input ports of the data acquisition module; The PWM control signal output terminal of the control logic module is connected to the H-bridge drive circuit input terminal of the execution drive module through a cable, and its I2C bus interface is connected to the RS-485 chip communication terminal of the communication interface module; The relay contact output end of the execution drive module is directly connected to the power control end of the sampling pump and the power interface of the solenoid valve of the pipeline; The 4G wireless unit of the communication interface module is connected to the microcontroller data exchange terminal of the control logic module through the UART serial port; The 12V DC output terminal of the power management module supplies power to the data acquisition module, the control logic module and the execution drive module through the power line; The relay control terminal of the feedback execution submodule is connected to the standby pump driving circuit of the execution driving module.

[0007] In a preferred embodiment, the data acquisition module is internally equipped with a high-precision electromagnetic flowmeter, a quartz crystal timer, a multi-parameter water quality sensor, and a signal conditioning circuit. The high-precision electromagnetic flowmeter uses a dual-electrode structure to monitor water flow velocity in the pipeline in real time through the Faraday electromagnetic induction principle, outputting an analog voltage signal. The quartz crystal timer has a built-in 32.768 kHz oscillator, providing a millisecond-level time reference signal. The multi-parameter water quality sensor integrates a turbidity probe, pH electrode, and conductivity measurement unit, supporting simultaneous acquisition of water quality parameters. The signal conditioning circuit includes a low-pass filter and an operational amplifier to eliminate high-frequency noise in the sensor signal and amplify weak electrical signals, ensuring stable data transmission to the control logic module.

[0008] In a preferred embodiment, the control logic module is internally provided with an STM32 series microcontroller, a PID closed-loop control algorithm, a non-volatile memory, and a real-time clock unit. The STM32 microcontroller receives flow sensor and time data through an SPI interface and parses the water quality sensor signal; the PID closed-loop control algorithm dynamically adjusts the PWM duty cycle of the sampling pump according to a preset flow threshold to achieve precise flow control; the non-volatile memory stores user-set sampling parameters and calibration data, and supports data retention after power failure; the real-time clock unit is synchronized with a quartz crystal timer, records sampling start and end timestamps, and communicates with the microcontroller via an I2C bus.

[0009] In a preferred embodiment, the execution drive module is internally equipped with an H-bridge motor driver chip, a diaphragm pump power control unit, a relay switching circuit, and a fault detection circuit. The H-bridge motor driver chip receives PWM signals from a microcontroller and outputs an adjustable voltage to drive the diaphragm pump motor in forward and reverse rotation. The diaphragm pump power control unit monitors the motor load in real time via a current feedback loop to prevent overcurrent or stalling. The relay switching circuit uses a magnetic latching relay to quickly disconnect and connect the main pump power supply. The fault detection circuit integrates a temperature sensor and voltage comparator to immediately trigger a protection mechanism when the motor temperature exceeds a safety threshold or the power supply voltage is abnormal.

[0010] In a preferred embodiment, the communication interface module is internally equipped with an RS-485 communication chip, a 4G wireless transmission unit, an LCD touchscreen interface, and a Modbus protocol stack. The RS-485 communication chip supports half-duplex communication and transmits data to the remote monitoring terminal via differential signals. The 4G wireless transmission unit has a built-in SIM card slot and baseband processor, which encapsulates sampled data in JSON format and uploads it to the cloud platform. The LCD touchscreen interface uses an RGB parallel bus to connect to a 128×64 pixel LCD screen, providing a local parameter setting interface. The Modbus protocol stack is embedded in the microcontroller, enabling standard data frame interaction with external devices.

[0011] In a preferred embodiment, the power management module is internally equipped with a lithium-ion battery pack, a DC-DC buck converter, an overvoltage protection circuit, and a low-power sleep unit. The lithium-ion battery pack uses a series-connected 18650 cell structure, providing a 12V DC output and supporting external power switching. The DC-DC buck converter stabilizes the battery voltage to 5V and 3.3V, which are supplied to the microcontroller and sensor, respectively. The overvoltage protection circuit uses a TVS diode and a resettable fuse to prevent lightning strikes and power surges. The low-power sleep unit shuts down non-core circuits during non-sampling periods, maintaining only the real-time clock and wake-up interrupt functions to maximize battery life.

[0012] In a preferred embodiment, the water quality data acquisition submodule is internally equipped with a multi-parameter water quality sensor, a signal conditioning circuit, an analog-to-digital conversion unit, and a communication interface protocol. The multi-parameter water quality sensor integrates a turbidity optical probe, a glass pH electrode, and a four-electrode conductivity measurement unit, acquiring water quality data based on scattered light intensity, potential difference, and current change. The signal conditioning circuit uses an LM358 operational amplifier to build a low-pass filter network to eliminate high-frequency interference noise in the sensor output signal. The analog-to-digital conversion unit is equipped with an ADS1115 high-precision 16-bit ADC chip, converting analog signals into digital quantities for transmission via the I2C bus. The communication interface protocol supports the RS-485 standard, and data is uploaded to the control logic module via a shielded twisted pair cable.

[0013] In a preferred embodiment, the abnormal analysis decision submodule realizes state discrimination based on multimodal feature fusion and time series pattern recognition. First, a state matrix containing acceleration frequency domain energy Ea, pressure gradient The 16-dimensional feature vector F∈R16 of the water level change rate H˙ and the temperature mutation ΔT is extracted using a sliding window mechanism to extract the joint features in the time and frequency domains; the feature compression is performed through an improved one-dimensional deep separable convolutional network to generate a low-dimensional latent vector z∈R 4 , synchronously calculate the KL divergence D of the real-time data stream and the historical normal pattern library KL ; Innovatively introduce dynamic confidence gating mechanism, trigger multi-expert voting decision when DKL>θ, and integrate vibration positioning deviation δ vib Matching residual with geomagnetic field mag Generate a comprehensive anomaly index A and finally update the decision boundary through online incremental learning; The calculation formula of the multimodal feature fusion function is: zi∈R4: latent feature vector of the i-th modality, extracted by a one-dimensional depthwise separable convolutional network; μi∈R: the mean value of the i-th modal feature in the historical normal data set, with a calculation period of 30 days; σi∈R + Indicates the standard deviation of the historical normal data set, updated daily; δvib∈[0,2.5] represents the positioning deviation output by the vibration wave positioning correction unit (unit: meter); ∈mag∈[0,1] represents the geomagnetic fingerprint matching residual, which is calculated by the improved DTW algorithm; αi∈R represents the learnable parameters, which are optimized by back-propagation; β = 0.35 represents the fusion weight of vibration positioning deviation, which is calibrated by 500 sets of measured data; The calculation formula of the dynamic confidence update rule is: Where: γ = 0.92 represents the historical confidence retention coefficient, corresponding to a half-life of approximately 8 hours; θt∈R+ represents the abnormality judgment threshold at time t, with an initial value of 0.7; N = 256 represents the sliding window capacity, corresponding to 5 minutes of data (sampling rate 1 Hz); I(·) represents the indicator function, which outputs 1 when An>θt, otherwise 0; An∈[0,5] represents the comprehensive anomaly index of the nth sample.

[0014] In a preferred embodiment, the feedback execution submodule is internally equipped with a magnetic latching relay array, an H-bridge motor drive circuit, an audible and visual alarm assembly, and an isolation protection circuit. The magnetic latching relay array uses a dual-contact structure, controlling the on / off power supply of the backup sampling pump via a coil drive circuit. The H-bridge motor drive circuit is based on an L298N chip design, receiving a microcontroller PWM signal and outputting a bidirectional current to drive the backup pump in both forward and reverse directions. The audible and visual alarm assembly includes a piezoelectric buzzer and a two-color LED light assembly with a frequency range of 2,000-4,000 Hz. The LED light assembly is connected to a 5V power supply via a current-limiting resistor. The isolation protection circuit integrates an optocoupler isolator and a TVS diode to block the reverse impact of high-voltage surges on the control unit.

[0015] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are: 1. The present invention improves sampling accuracy and operational efficiency through an automated control unit and modular design, reducing human error. The device automatically completes timed and quantitative extraction, enhancing real-time monitoring of groundwater flow and quality, ensuring accurate and reliable sampling data. Furthermore, corrosion-resistant materials and a low-power power supply extend the device's service life in complex environments. The modular architecture facilitates expansion of water quality parameter detection capabilities to meet diverse needs in different scenarios, reducing maintenance costs and operational complexity.

[0016] 2. The present invention significantly enhances the intelligence and data reliability of the sampling process by integrating a water quality adaptive module. This module monitors key parameters such as turbidity and conductivity in groundwater in real time and automatically triggers secondary sampling or an alarm mechanism when an anomaly is detected. This design avoids the problem of sample invalidation caused by sudden changes in water quality in traditional sampling, ensuring that collected water samples meet monitoring standards. It also reduces the frequency of manual re-debugging and on-site maintenance, improving operational efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 Schematic diagram of the device structure of the present invention; Figure 2 This is a system block diagram of the control unit in the present invention; Figure 3 This is a system block diagram of the water quality adaptive module in the present invention.

[0018] Markings in the figure: 1-sampling pump, 2-pipeline, 3-flow sensor, 4-water storage container, 5-timer, 6-control unit, 7-data acquisition module, 8-control logic module, 9-execution drive module, 10-communication interface module, 11-power management module, 12-water quality adaptive module, 13-water quality data acquisition submodule, 14-abnormal analysis and decision submodule, 15-feedback execution submodule. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0020] Example: Reference Figure 1-3 , a groundwater extraction sampling device, the sampling device includes a sampling pump 1, a pipeline 2, a flow sensor 3, a water storage container 4, a timer 5 and a control unit 6; The water outlet of the sampling pump 1 is connected to the groundwater well through the inlet end of the stainless steel corrugated pipe 2, and the outlet end of the pipe 2 is connected to the water inlet of the water storage container 4; The flow sensor 3 is embedded in the middle section of the pipeline 2, and the electrode signal output end of the flow sensor 3 is connected to the input end of the data acquisition module 7 of the control unit 6 through a shielded cable; The liquid level float switch signal line of the water storage container 4 is connected to the input port of the logic control module 8 of the control unit 6. The clock signal output terminal of the timer 5 communicates with the microcontroller of the control unit 6 via the I2C bus; The PWM signal output terminal of the execution drive module 9 of the control unit 6 is connected to the motor drive circuit of the sampling pump 1; The control unit 6 is internally provided with a data acquisition module 7, a control logic module 8, an execution drive module 9, a communication interface module 10, a power management module 11 and a water quality adaptation module 12; The water quality self-adaptation module 12 is composed of a water quality data acquisition submodule 13, an abnormality analysis and decision submodule 14, and a feedback execution submodule 15. The sensor signal output end of the data acquisition module 7 is connected to the SPI communication input end of the control logic module 8 through a shielded cable, wherein the electrode signal of the flow sensor 3 and the clock signal of the timer 5 are respectively connected to the analog and digital input ports of the data acquisition module 7; The PWM control signal output terminal of the control logic module 8 is connected to the H-bridge drive circuit input terminal of the execution drive module 9 via a cable, and its I2C bus interface is connected to the RS-485 chip communication terminal of the communication interface module (10); The relay contact output terminal of the execution drive module 9 is directly connected to the power control terminal of the sampling pump 1 and the solenoid valve power interface of the pipeline 2; The 4G wireless unit of the communication interface module 10 is connected to the data exchange terminal of the microcontroller of the control logic module 8 via the UART serial port; The 12V DC output terminal of the power management module 11 supplies power to the data acquisition module 7, the control logic module 8 and the execution drive module 9 through the power line; The relay control terminal of the feedback execution submodule 15 is connected to the standby pump drive circuit of the execution drive module 9 .

[0021] Data acquisition module 7 houses a high-precision electromagnetic flowmeter, a quartz crystal timer, a multi-parameter water quality sensor, and a signal conditioning circuit. The high-precision electromagnetic flowmeter utilizes a dual-electrode structure, employing the principle of Faraday electromagnetic induction to monitor water flow velocity in the pipeline in real time and output an analog voltage signal. The quartz crystal timer includes a built-in 32.768 kHz oscillator, providing a millisecond-level time reference signal. The multi-parameter water quality sensor integrates a turbidity probe, pH electrode, and conductivity measurement unit, enabling simultaneous acquisition of water quality parameters. The signal conditioning circuit, comprised of a low-pass filter and operational amplifier, eliminates high-frequency noise in sensor signals and amplifies weak electrical signals, ensuring stable data transmission to the control logic module.

[0022] Control logic module 8 incorporates an STM32 microcontroller, a PID closed-loop control algorithm, non-volatile memory, and a real-time clock unit. The STM32 microcontroller receives flow sensor and time data via the SPI interface and analyzes the water quality sensor signal. The PID closed-loop control algorithm dynamically adjusts the sampling pump's PWM duty cycle based on a preset flow threshold to achieve precise flow control. The non-volatile memory stores user-defined sampling parameters and calibration data, retaining data even after power failure. The real-time clock unit, synchronized with a quartz crystal timer, records sampling start and end timestamps and communicates with the microcontroller via the I2C bus.

[0023] The execution drive module 9 houses an H-bridge motor driver chip, a diaphragm pump power control unit, a relay switching circuit, and a fault detection circuit. The H-bridge motor driver chip receives PWM signals from the microcontroller and outputs an adjustable voltage to drive the diaphragm pump motor in both forward and reverse directions. The diaphragm pump power control unit monitors the motor load in real time via a current feedback loop to prevent overcurrent or stalling. The relay switching circuit uses a magnetic latching relay to quickly disconnect and connect the main pump power supply. The fault detection circuit integrates a temperature sensor and voltage comparator to trigger protection mechanisms immediately when the motor temperature exceeds a safety threshold or the power supply voltage is abnormal.

[0024] The communication interface module 10 is internally equipped with an RS-485 communication chip, a 4G wireless transmission unit, an LCD touchscreen interface, and a Modbus protocol stack. The RS-485 communication chip supports half-duplex communication and transmits data to the remote monitoring terminal via differential signals. The 4G wireless transmission unit has a built-in SIM card slot and baseband processor, which encapsulates sampled data in JSON format and uploads it to the cloud platform. The LCD touchscreen interface uses an RGB parallel bus to connect to a 128×64 pixel LCD screen, providing a local parameter setting interface. The Modbus protocol stack is embedded in the microcontroller, enabling standard data frame interaction with external devices.

[0025] The power management module 11 houses a lithium-ion battery pack, a DC-DC buck converter, an overvoltage protection circuit, and a low-power sleep unit. The lithium-ion battery pack utilizes 18650 cells in series, providing a 12V DC output and supporting external power switching. The DC-DC buck converter stabilizes the battery voltage to 5V and 3.3V, supplying the microcontroller and sensor, respectively. The overvoltage protection circuit uses a TVS diode and a resettable fuse to protect against lightning strikes and power surges. The low-power sleep unit shuts down non-core circuits during non-sampling periods, maintaining only the real-time clock and wake-up interrupt functions to maximize battery life.

[0026] The water quality data acquisition submodule 13 houses a multi-parameter water quality sensor, signal conditioning circuitry, an analog-to-digital conversion unit, and a communication interface protocol. The multi-parameter water quality sensor integrates a turbidity optical probe, a glass pH electrode, and a four-electrode conductivity measurement unit. It acquires water quality data based on scattered light intensity, potential difference, and current changes, respectively. The signal conditioning circuit utilizes an LM358 operational amplifier to form a low-pass filter network, eliminating high-frequency interference noise in the sensor output signal. The analog-to-digital conversion unit, equipped with an ADS1115 high-precision 16-bit ADC chip, converts analog signals into digital quantities and transmits them via the I2C bus. The communication interface protocol supports the RS-485 standard, and data is uploaded to the control logic module via a shielded twisted-pair cable.

[0027] The abnormal analysis decision submodule 14 realizes state discrimination based on multimodal feature fusion and time series pattern recognition. First, it constructs a model including acceleration frequency domain energy Ea, pressure gradient The 16-dimensional feature vector F∈R16 of the water level change rate H˙ and the temperature mutation ΔT is extracted using a sliding window mechanism to extract the joint features in the time and frequency domains; the feature compression is performed through an improved one-dimensional deep separable convolutional network to generate a low-dimensional latent vector z∈R 4 , synchronously calculate the KL divergence D of the real-time data stream and the historical normal pattern library KL ; Innovatively introduce dynamic confidence gating mechanism, trigger multi-expert voting decision when DKL>θ, and integrate vibration positioning deviation δ vib Matching residual with geomagnetic field mag Generate a comprehensive anomaly index A and finally update the decision boundary through online incremental learning; The calculation formula of the multimodal feature fusion function is: zi∈R4: latent feature vector of the i-th modality, extracted by a one-dimensional depthwise separable convolutional network; μi∈R: the mean value of the i-th modal feature in the historical normal data set, with a calculation period of 30 days; σi∈R + Indicates the standard deviation of the historical normal data set, updated daily; δvib∈[0,2.5] represents the positioning deviation output by the vibration wave positioning correction unit (unit: meter); ∈mag∈[0,1] represents the geomagnetic fingerprint matching residual, which is calculated by the improved DTW algorithm; αi∈R represents the learnable parameters, which are optimized by back-propagation; β = 0.35 represents the fusion weight of vibration positioning deviation, which is calibrated by 500 sets of measured data; The calculation formula of the dynamic confidence update rule is: Where: γ = 0.92 represents the historical confidence retention coefficient, corresponding to a half-life of approximately 8 hours; θt∈R+ represents the abnormality judgment threshold at time t, with an initial value of 0.7; N = 256 represents the sliding window capacity, corresponding to 5 minutes of data (sampling rate 1 Hz); I(·) represents the indicator function, which outputs 1 when An>θt, otherwise 0; An∈[0,5] represents the comprehensive anomaly index of the nth sample; The feedback execution submodule 15 is internally equipped with a magnetic latching relay array, an H-bridge motor drive circuit, an audible and visual alarm assembly, and an isolation protection circuit. The magnetic latching relay array uses a dual-contact structure, controlling the on / off power supply of the backup sampling pump via a coil drive circuit. The H-bridge motor drive circuit is based on an L298N chip design, receiving PWM signals from a microcontroller and outputting bidirectional current to drive the backup pump in both forward and reverse directions. The audible and visual alarm assembly includes a piezoelectric buzzer and a two-color LED light assembly with a frequency range of 2,000 to 4,000 Hz. The LED light assembly is connected to a 5V power supply via a current-limiting resistor. The isolation protection circuit integrates an optocoupler and a TVS diode to block high-voltage surges from impacting the control unit.

[0028] From the above we can know: This invention improves sampling accuracy and operational efficiency through an automated control unit and modular design, reducing human error. The device automatically performs timed and quantitative extraction, enhancing real-time monitoring of groundwater flow and quality, ensuring accurate and reliable sampling data. Furthermore, corrosion-resistant materials and a low-power power supply extend the device's service life in complex environments. The modular architecture facilitates expansion of water quality parameter detection capabilities to meet diverse needs in different scenarios, reducing maintenance costs and operational complexity.

[0029] The present invention significantly enhances the intelligence and data reliability of the sampling process by integrating a water quality adaptive module. This module monitors key parameters such as turbidity and conductivity in groundwater in real time and automatically triggers secondary sampling or an alarm mechanism when anomalies are detected. This design avoids the problem of sample invalidation caused by sudden changes in water quality in traditional sampling, ensuring that collected water samples meet monitoring standards. It also reduces the frequency of manual re-testing and on-site maintenance, improving operational efficiency.

[0030] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprises" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device that includes a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further limitations, an element defined by the sentence "comprises a ..." does not exclude the presence of other identical elements in the process, method, article or device that includes the element.

[0031] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A groundwater extraction and sampling device, characterized in that: The sampling device comprises a sampling pump (1), a pipeline (2), a flow sensor (3), a water storage container (4), a timer (5) and a control unit (6); The water outlet of the sampling pump (1) is connected to the groundwater well through the inlet end of the stainless steel corrugated pipe (2), and the outlet end of the pipe (2) is connected to the water inlet of the water storage container (4); The flow sensor (3) is embedded in the middle section of the pipeline (2), and the electrode signal output end of the flow sensor (3) is connected to the input end of the data acquisition module (7) of the control unit (6) through a shielded cable; The liquid level float switch signal line of the water storage container (4) is connected to the input port of the logic control module (8) of the control unit (6). The clock signal output terminal of the timer (5) communicates with the microcontroller of the control unit (6) via an I2C bus; The PWM signal output end of the execution drive module (9) of the control unit (6) is connected to the motor drive circuit of the sampling pump (1); The control unit (6) is internally provided with a data acquisition module (7), a control logic module (8), an execution drive module (9), a communication interface module (10), a power management module (11) and a water quality self-adaptation module (12); The water quality self-adaptation module (12) is composed of a water quality data acquisition submodule (13), an abnormality analysis and decision submodule (14), and a feedback execution submodule (15).

2. A groundwater extraction and sampling device according to claim 1, characterized in that: The sensor signal output end of the data acquisition module (7) is connected to the SPI communication input end of the control logic module (8) via a shielded cable, wherein the electrode signal of the flow sensor (3) and the clock signal of the timer (5) are respectively connected to the analog and digital input ports of the data acquisition module (7); The PWM control signal output terminal of the control logic module (8) is connected to the H-bridge drive circuit input terminal of the execution drive module (9) via a cable, and its I2C bus interface is connected to the RS-485 chip communication terminal of the communication interface module (10); The relay contact output end of the execution drive module (9) is directly connected to the power control end of the sampling pump (1) and the solenoid valve power interface of the pipeline (2); The 4G wireless unit of the communication interface module (10) is connected to the microcontroller data exchange terminal of the control logic module (8) via a UART serial port; The 12V DC output terminal of the power management module (11) supplies power to the data acquisition module (7), the control logic module (8) and the execution drive module (9) via a power line; The relay control terminal of the feedback execution submodule (15) is connected to the standby pump drive circuit of the execution drive module (9).

3. A groundwater extraction and sampling device according to claim 1, characterized in that: The data acquisition module (7) is internally provided with a high-precision electromagnetic flowmeter, a quartz crystal timer, a multi-parameter water quality sensor and a signal conditioning circuit.

4. A groundwater extraction and sampling device according to claim 1, characterized in that: The control logic module (8) is internally provided with an STM32 series microcontroller, a PID closed-loop control algorithm, a non-volatile memory and a real-time clock unit; the STM32 microcontroller receives flow sensor and time data through an SPI interface and analyzes water quality sensor signals; The PID closed-loop control algorithm dynamically adjusts the sampling pump's PWM duty cycle based on the preset flow threshold to achieve precise flow control. The non-volatile memory stores user-set sampling parameters and calibration data, and supports data retention after power failure. The real-time clock unit is synchronized with the quartz crystal timer, records the sampling start and end timestamps, and communicates with the microcontroller through the I2C bus.

5. A groundwater extraction and sampling device according to claim 1, characterized in that: The execution drive module (9) is internally provided with an H-bridge motor drive chip, a diaphragm pump power control unit, a relay switch circuit and a fault detection circuit.

6. A groundwater extraction and sampling device according to claim 1, characterized in that: The communication interface module (10) is internally provided with an RS-485 communication chip, a 4G wireless transmission unit, an LCD touch screen interface and a Modbus protocol stack.

7. A groundwater extraction and sampling device according to claim 1, characterized in that: The power management module (11) is internally provided with a lithium-ion battery pack, a DC-DC buck converter, an overvoltage protection circuit and a low-power sleep unit.

8. A groundwater extraction and sampling device according to claim 1, characterized in that: The water quality data acquisition submodule (13) is internally provided with a multi-parameter water quality sensor, a signal conditioning circuit, an analog-to-digital conversion unit and a communication interface protocol.

9. A groundwater extraction and sampling device according to claim 1, characterized in that: The abnormal analysis decision submodule (14) realizes state discrimination based on multimodal feature fusion and time series pattern recognition. First, a state matrix containing acceleration frequency domain energy Ea, pressure gradient The 16-dimensional feature vector F∈R16 of the water level change rate H˙ and the temperature mutation ΔT is extracted using a sliding window mechanism to extract the joint features in the time and frequency domains; Feature compression is performed through an improved one-dimensional depth-wise separable convolutional network to generate a low-dimensional latent vector z∈R 4 , synchronously calculate the KL divergence D of the real-time data stream and the historical normal pattern library KL ; Innovatively introduce dynamic confidence gating mechanism, when D KL >θ triggers multi-expert voting decision and integrates vibration positioning deviation δ vib Matching residual with geomagnetic field mag Generate a comprehensive anomaly index A and finally update the decision boundary through online incremental learning; The calculation formula of the multimodal feature fusion function is: zi∈R4: latent feature vector of the i-th modality, extracted by a one-dimensional depthwise separable convolutional network; μi∈R: the mean value of the i-th modal feature in the historical normal data set, with a calculation period of 30 days; σi∈R + Indicates the standard deviation of the historical normal data set, updated daily; δvib∈[0,2.5] represents the positioning deviation output by the vibration wave positioning correction unit; ∈mag∈[0,1] represents the geomagnetic fingerprint matching residual, which is calculated by the improved DTW algorithm; αi∈R represents the learnable parameters, which are optimized by back-propagation; β = 0.35 represents the fusion weight of vibration positioning deviation, which is calibrated by 500 sets of measured data; The calculation formula of the dynamic confidence update rule is: Where: γ = 0.92 represents the historical confidence retention coefficient, corresponding to a half-life of approximately 8 hours; θt∈R+ represents the abnormality judgment threshold at time t, with an initial value of 0.7; N = 256 represents the sliding window capacity, corresponding to 5 minutes of data; I(·) represents the indicator function, which outputs 1 when An>θt, otherwise 0; An∈[0,5] represents the comprehensive anomaly index of the nth sample.

10. The groundwater extraction and sampling device according to claim 1, characterized in that: The feedback execution submodule (15) is internally provided with a magnetic latching relay array, an H-bridge motor drive circuit, an audible and visual alarm component, and an isolation protection circuit.

Images