A sampling device and sampling method for water environment monitoring

Abstract

This invention relates to the field of environmental monitoring technology and discloses a sampling device and method for water environment monitoring, comprising: pre-certifying sensor reliability, performing pre-inspection using a biomimetic sampling head, calculating parameter decision weights to quantify sensor reliability, establishing a travel time model, calculating the travel time of the sampling device and planning a dynamic path for sampling based on the new sampling course; performing laminar flow control sampling using countercurrent shear stratification technology, calculating the critical settling velocity to achieve zero mixing between layers, simultaneously performing on-chip fixed decision-making, and constructing a segmented model by calculating the correction signal intensity to inversely extrapolate the concentration value of the detected object; correcting abnormal federated values ​​based on the segmented model using a loss function and performing global aggregation, performing quantum evidence storage based on the aggregation completion, and simultaneously establishing a self-healing sealed chamber and a dynamic temperature-controlled sample library to achieve non-destructive preservation of sample activity.

Description

Technical Field

[0001] This invention relates to the field of environmental monitoring technology, and specifically to a sampling device and sampling method for water environment monitoring. Background Technology

[0002] Traditional sampling devices cannot achieve accurate sampling of stratified water bodies. In shallow clay layers in karst areas, clay heads are easily disturbed by soil clumps and gravel, leading to sampling interruptions. Although some patents propose stratified sampling structures, they require frequent adjustments to equipment parameters, making operation complex and unable to dynamically adapt to changes in water stratification. During long-term monitoring, the sensor surface is easily covered by biofilm, and existing technologies lack long-term anti-adhesion designs, resulting in distorted detection data.

[0003] In existing technologies, reliance on manual operation or preset programs makes it impossible to dynamically adjust sampling strategies based on real-time monitoring data, and remote control response delays or data transmission interruptions are common problems. The use of independent sensors to collect different parameters lacks a high-precision synchronization mechanism, resulting in weak data correlation. Multi-parameter sampling in the aquatic environment may introduce system biases due to inaccurate calibration. Traditional samplers are difficult to install and prone to clogging; in aquatic environments, both corrosion resistance and biocompatibility must be considered, which existing patents have not adequately addressed. Traditional stratified monitoring wells pose a risk of interconnection between upper and lower layers, and the well-washing process is complex, failing to guarantee sample representativeness. The lack of sufficient integration of microfluidics and Raman spectroscopy in aquatic environment sampling leads to a lack of rapid on-site analysis capabilities. Existing patents lack machine learning-based anomaly data identification models, preventing intelligent decision-making like that of truck coal sampling systems.

[0004] Therefore, there is a need to provide a sampling device and sampling method for water environment monitoring. Summary of the Invention

[0005] The purpose of this invention is to provide a sampling device and sampling method for water environment monitoring. To solve the above-mentioned problems in the prior art, this invention achieves this through the following technical solution:

[0006] In a first aspect, the present invention provides a sampling device and sampling method for water environment monitoring, which specifically includes the following steps:

[0007] Step 1: Introduce a digital twin pre-inspection system and a quantum-encrypted blockchain calibration chain to pre-certify the trustworthiness of the sensor. Perform pre-inspection through a bionic sampling head, obtain the standard deviation of the sensor's historical error, calculate the parameter decision weight to quantify the sensor's trustworthiness, obtain the target point coordinates to establish a flight time model, calculate the flight time of the sampling device and the sampling new course to plan a dynamic path for sampling.

[0008] Step 2: Based on the planned dynamic path, laminar flow control sampling is performed using counter-current shear stratification technology. The critical settling velocity is calculated to achieve zero mixing between layers. On-chip fixed decision is performed simultaneously based on laminar flow control sampling. A segmented model is constructed by calculating the correction signal strength to inversely calculate the concentration value of the detected object.

[0009] Step 3: Acquire and fuse multi-source data, correct abnormal federated values ​​based on a segmented model using a loss function and perform global aggregation, perform quantum evidence storage based on the aggregation completion, and simultaneously establish a self-healing sealed chamber and a dynamic temperature-controlled sample library to achieve non-destructive preservation of sample activity.

[0010] As a further invention of this solution, the method for performing the pre-detection is as follows:

[0011] Pre-testing is performed using a biomimetic sampling head. Test solution containing fluorescent microspheres is introduced into the sampling head. The microspheres are made of polystyrene, and the microsphere throughput is counted in real time.

[0012] Calculate smoothness ,in This represents the theoretical throughput of microspheres. To test the fluid flow rate, a peristaltic pump was used for precise control. This represents the total area of ​​the gill openings. This represents the actual amount of microspheres passing through.

[0013] like If the result is greater than 95%, the system will automatically start ultrasonic cleaning and repeat the test until the standard is met.

[0014] The monthly average of pollution data for the sampling area over the past three years is obtained. The sensor calibration certificate is retrieved via blockchain and verified using the SHA-256 hash algorithm. If the hash value matches the manufacturer's database, the verification is successful.

[0015] As a further invention of this solution, the method for calculating the decision weights of the parameters is as follows:

[0016] The historical error standard deviation of the sensor can be obtained using the formula:

[0017]

[0018] Calculate the parameter decision weights ,in and The standard deviation of the sensor's historical error. and This represents the sensor's annual drift rate; if the sensor's historical error standard deviation is greater than 0.1, it will be automatically marked as standby.

[0019] A four-dimensional hydrodynamic positioning algorithm is adopted to fuse flow velocity vectors and turbulence intensity, replacing the traditional coarse GPS positioning.

[0020] As a further invention of this solution, the method for calculating the navigation time and sampling the new heading is as follows:

[0021] Get the coordinates of the target point Sampling is performed by planning dynamic paths;

[0022] Obtain and input the target point coordinates The three-dimensional flow field is acquired in real time using an ADCP flowmeter. ;

[0023] A flight time model is established based on the target point coordinates, using the formula:

[0024]

[0025] Calculate the sailing time ,in, The straight-line distance to the target point and , The mean velocity vector magnitude, The intensity of turbulent fluctuations is calculated using the standard deviation of the flow velocity. The angle between the direction of the water flow and the navigation path. It has a moderate turbulence coefficient. This is the maximum acceleration of the thruster;

[0026] Based on the calculated travel time, an underwater digital elevation model (DEM) is generated using multibeam sonar for terrain-adaptive obstacle avoidance.

[0027] If the measured distance d is less than the preset safe distance threshold, the three-dimensional coordinates of the obstacle surface points are determined by multibeam sonar, and the obstacle normal vector is obtained by extracting the unit vector of the surface normal through geometric fitting. Based on the obstacle normal vector, using the formula:

[0028]

[0029] Calculate the new heading for sampling ,in, Original heading The roughness coefficient is the obstacle surface roughness coefficient. The obstacle normal vector;

[0030] As a further invention of this solution, the method for calculating the critical settling velocity is as follows:

[0031] Laminar flow control sampling is performed based on temperature sensor array data, and the target depth is determined by combining the thermocline algorithm, using the formula:

[0032]

[0033] The critical settling velocity was calculated. ,in, It is the acceleration due to gravity. For particulate matter density, For water density, For particulate matter particle size, This is the drag coefficient;

[0034] Based on the obtained critical settling velocity, the sampling head is adjusted by a vector thruster to cut in at a 15° countercurrent angle, dynamically adjusting the gill opening area, and using a Doppler current meter for real-time feedback adjustment to stabilize the inflow velocity within a preset range.

[0035] As a further invention of this solution, the method for making on-chip fixed decisions is as follows:

[0036] For heavy metal detection: When the ORP sensor reading is less than the preset reading threshold, the microfluidic chip automatically injects 65% pure AR-grade HNO3, and the injection volume is adjusted by the pH sensor feedback, with the target pH value being less than 2.

[0037] For volatile organic compounds: When the temperature sensor detects that the temperature is higher than the preset temperature threshold, the Peltier element starts cooling, which lowers the sample chamber to the preset cooling temperature within the preset cooling cycle and triggers the electromagnetic sealing valve to close.

[0038] For sulfides: When the DO sensor reading is less than the sulfide content threshold, a Zn(Ac)2 solution with a preset volume ratio is injected to generate ZnS precipitate;

[0039] The dual-laser SERS technology enables full-spectrum detection within a preset frequency band, covering the blind spots of traditional methods.

[0040] As a further invention of this scheme, the method for performing concentration inversion is as follows:

[0041] Turbidity of microplastic samples was obtained using online sensors. The original sensor signal intensity is corrected based on the turbidity of the microplastic sample, using the correction formula:

[0042]

[0043] The corrected signal strength is calculated. ,in, The original sensing signal strength, It is a natural constant. For the water temperature of the microplastic sample, Turbidity of microplastic samples This is the temperature influence coefficient;

[0044] Concentration inversion is performed by constructing a segmented model based on the strength of the corrected signal;

[0045] As a further invention of this solution, the method for correcting and globally aggregating abnormal federated values ​​is as follows:

[0046] SERS spectra, hydrological parameters, and in-situ images were acquired, and the sub-indices were normalized and analyzed to obtain sub-indices. The comprehensive index was then calculated using a weighted fusion formula.

[0047] Based on a comprehensive index combined with a local model, outlier federated values ​​are corrected and globally aggregated using a loss function;

[0048] Based on the local model, using the loss function:

[0049]

[0050] Correcting abnormal federated values, among which, For model parameters, For sub-index weights, These are measured values. These are the model's predicted values. The residual is the measured value.

[0051] As a further invention of this solution, the method for the dynamic temperature-controlled sample library is as follows:

[0052] Constructing a dynamic temperature-controlled sample library: Utilizing a temperature control algorithm, through the formula:

[0053]

[0054] The calculated sample library control temperature ,in, The optimal storage temperature for contaminants, Correction factor for container heat capacity , For ambient temperature, The temperature coefficient of degradation rate, For estimated storage time;

[0055] The phase change material chamber maintains a preset range and periodically records the temperature data of the sample library through blockchain within a preset recording period, forming an immutable temperature traceability chain.

[0056] Secondly, the sampling system for water environment monitoring provided in this embodiment of the invention specifically includes the following modules:

[0057] Pre-inspection planning module: Introduces a digital twin pre-inspection system and a quantum encrypted blockchain calibration chain to pre-certify the trustworthiness of sensors. Pre-inspection is carried out through a bionic sampling head to obtain the historical error standard deviation of sensors, calculate the parameter decision weight to quantify sensor trustworthiness, obtain the target point coordinates to establish a flight time model, calculate the flight time of the sampling device and the sampling new course to plan a dynamic path for sampling;

[0058] Layered fixing module: Based on the planned dynamic path, it uses counter-current shearing layering technology for laminar flow control sampling, calculates the critical settling velocity to achieve zero mixing between layers, performs on-chip fixing decision based on laminar flow control sampling, and constructs a segmented model by calculating the correction signal strength to invert the concentration value of the detected object.

[0059] The fusion and evidence storage module acquires and fuses multi-source data, corrects abnormal federated values ​​based on a segmented model using a loss function, and performs global aggregation. Quantum evidence storage is then performed based on the aggregation, and a self-healing sealed chamber and a dynamic temperature-controlled sample library are established simultaneously to achieve non-destructive preservation of sample activity.

[0060] Thirdly, the sampling device for water environment monitoring provided in the embodiments of the present invention specifically includes the following:

[0061] This device includes: a digital twin module, a quantum encrypted blockchain module, and a dual-laser SERS detection unit; the sampling head adopts a biomimetic gill structure with three symmetrically distributed gill lobes, a single-lobe opening area, and built-in fluorescent microsphere detection channels and self-cleaning components; the intelligent control system integrates a federated learning framework and a four-dimensional hydrodynamic algorithm; the sample storage chamber is upgraded to a self-healing sealed structure and equipped with a phase change material temperature control system; the power system uses a vector thruster, supports heading adjustment, and meets the requirements of countercurrent stratified sampling.

[0062] The beneficial effects of this invention are:

[0063] 1. Real-time cloud mapping of component status is achieved through digital twins, combined with a sensor calibration chain based on quantum-encrypted blockchain, ensuring the immutability of historical sensor data; abandoning traditional GPS coarse positioning, it integrates ADCP three-dimensional flow field, turbulence intensity, and multibeam sonar DEM, calculating new course through flight time model and obstacle normal vector, improving path planning accuracy and adapting to complex underwater terrain; adopting a biomimetic sampling head with gill-like structure, it uses fluorescent microsphere test liquid to detect microsphere throughput in real time, combined with ultrasonic cleaning technology to achieve automatic cleaning; at the same time, it uses a laser scattering instrument to quantitatively evaluate the sampling head's smoothness, ensuring that the cleaning effect meets the standards; integrating SERS spectrum, hydrological parameters, and in-situ images, it uses analytic hierarchy process weighted sum loss function to correct outliers, improving the accuracy of the comprehensive index;

[0064] 2. Based on temperature sensor array data, a counter-current shear stratification technique was used to achieve zero mixing between layers. The critical settling velocity was calculated using a formula to solve the cross-contamination problem of traditional stratified sampling and to verify the feasibility of shear flow control. A signal correction formula based on turbidity and water temperature was established, and the problem of light scattering interference in complex water bodies was solved through standard sample calibration. A four-dimensional hydrodynamic positioning algorithm was used to integrate BeiDou positioning, ADCP velocity field data and turbulence intensity parameters to construct a time-of-flight model, achieving an accuracy that traditional GPS coarse positioning could not reach. Attached Figure Description

[0065] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying 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.

[0066] Figure 1 This is a flowchart of the sampling method for water environment detection provided in Embodiment 1 of the present invention;

[0067] Figure 2 This is a schematic diagram of the structure of a sampling system for water environment detection provided in Embodiment 2 of the present invention. Detailed Implementation

[0068] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0069] Example 1: As Figure 1 As shown in the figure, the sampling method for water environment detection provided by this embodiment of the invention specifically includes the following steps:

[0070] Step 1: Introduce a digital twin pre-inspection system and a quantum-encrypted blockchain calibration chain to pre-certify the trustworthiness of the sensor. Perform pre-inspection through a bionic sampling head, obtain the standard deviation of the sensor's historical error, calculate the parameter decision weight to quantify the sensor's trustworthiness, obtain the target point coordinates to establish a flight time model, calculate the flight time of the sampling device and the sampling new course to plan a dynamic path for sampling.

[0071] In a specific embodiment, a digital twin pre-inspection system and a quantum encrypted blockchain calibration chain are introduced into the sampling device to realize cloud mapping of component status and pre-certification of sensor credibility, replacing the traditional manual inspection mode;

[0072] Pre-testing is performed using a biomimetic sampling head. Test solution containing fluorescent microspheres is introduced into the sampling head. The microspheres are made of polystyrene, with a preset fluorescence wavelength of 520nm. The preset wavelength of the laser scattering instrument is 635nm, and the preset power is 10mW. The microsphere throughput is counted in real time.

[0073] Calculate smoothness ,in This represents the theoretical throughput of microspheres. To test the fluid flow rate, a peristaltic pump was used for precise control. This represents the total area of ​​the gill openings. This represents the actual amount of microspheres passing through.

[0074] like If the result is greater than 95%, the system will automatically start 40kHz ultrasonic cleaning for 10 minutes, with the power preset to 50W. Repeat the test until the standard is met.

[0075] For example, the measured Qactual = 72 μL / s, which, converted using microsphere counting, translates to 1.44 × 102 microspheres detected per second. 4 Each microsphere, conversion factor 5×10 5 Cells / mL = 5 × 10 2 cells / μL, Q theory = 3cm / s × 25mm 2 =30mm / s×25mm 2 =750mm 3 / s=75μL / s, η=72 / 75×100%=96%, qualified;

[0076] The monthly average of pollution data for the past three years in the sampling area was obtained through NASAEarthdata. The pollution data includes, but is not limited to, pH value, DO parameter and turbidity. The sensor calibration certificate was retrieved through blockchain and verified using SHA-256 hash algorithm. If the hash value matches the manufacturer's database, it passes the verification.

[0077] The historical error standard deviation of the sensor can be obtained using the formula:

[0078]

[0079] Calculate the parameter decision weights ,in and The standard deviation of the sensor's historical error is calculated based on the most recent 100 ISO 17025 certification calibration data. and The sensor's annual drift rate reflects its long-term stability degradation; if the sensor's historical error standard deviation is greater than 0.1, it is automatically marked as standby.

[0080] A four-dimensional hydrodynamic positioning algorithm is adopted to fuse flow velocity vectors and turbulence intensity, replacing the traditional coarse GPS positioning.

[0081] Get the coordinates of the target point Sampling is performed by planning dynamic paths;

[0082] Specifically, the target point coordinates are obtained and input through BeiDou positioning. The three-dimensional flow field is acquired in real time using an ADCP flowmeter. The data update frequency is preset to 1Hz;

[0083] A flight time model is established based on the target point coordinates, using the formula:

[0084]

[0085] Calculate the sailing time ,in, The straight-line distance to the target point and , The mean velocity vector magnitude, The intensity of turbulent fluctuations is calculated using the standard deviation of the flow velocity. The angle between the direction of the water flow and the navigation path. The turbulence coefficient is set to a moderate level, with a preset value of 0.3, and is dynamically adjusted based on the on-site velocity gradient. This is the maximum acceleration of the thruster;

[0086] Based on the calculated travel time, a 1cm resolution underwater digital elevation model (DEM) is generated using multibeam sonar for terrain-adaptive obstacle avoidance.

[0087] Specifically, if the measured distance d is less than the preset safe distance threshold, the three-dimensional coordinates of the obstacle surface points are determined by multibeam sonar, and the obstacle normal vector is obtained by extracting the unit vector of the surface normal through geometric fitting. Based on the obstacle normal vector, using the formula:

[0088]

[0089] Calculate the new heading for sampling ,in, Original heading The obstacle roughness coefficient is determined by DEM texture detected by multibeam sonar. The obstacle normal vector;

[0090] Step 2: Based on the planned dynamic path, laminar flow control sampling is performed using counter-current shear stratification technology. The critical settling velocity is calculated to achieve zero mixing between layers. On-chip fixed decision is performed simultaneously based on laminar flow control sampling. A segmented model is constructed by calculating the correction signal strength to inversely calculate the concentration value of the detected object.

[0091] By utilizing countercurrent shear stratification technology, zero mixing between layers is achieved, and in-situ chemical fixation is completed simultaneously, thus solving the problem of cross-contamination in traditional stratified sampling.

[0092] Specifically, laminar flow control sampling is performed based on temperature sensor array data, and the target depth is determined by combining the thermocline algorithm, using the formula:

[0093]

[0094] The critical settling velocity was calculated. ,in, It is the acceleration due to gravity. For particulate matter density, For water density, For particulate matter particle size, The drag coefficient is determined based on the Reynolds number. Segmented calculation, When less than or equal to 1, , When greater than 1 and less than 1000, , When greater than or equal to 1000, ;

[0095] Based on the obtained critical settling velocity, the sampling head is adjusted by the vector thruster to cut in at a 15° countercurrent angle, dynamically adjusting the gill opening area, and using the Doppler current meter to provide real-time feedback adjustment to stabilize the inflow velocity within the preset range, thus avoiding interlayer disturbance.

[0096] On-chip fixed decision-making is based on laminar flow control sampling synchronization;

[0097] For heavy metal detection: When the ORP sensor reading is less than the preset reading threshold, the microfluidic chip automatically injects 65% pure AR-grade HNO3, and the injection volume is adjusted by the pH sensor feedback, with the target pH value being less than 2.

[0098] For volatile organic compounds: When the temperature sensor detects that the temperature is higher than the preset temperature threshold, the Peltier element starts cooling with a preset power of 50W. Within the preset cooling cycle, the sample chamber is reduced to the preset cooling temperature and the electromagnetic sealing valve is triggered to close.

[0099] For sulfides: When the DO sensor reading is less than the sulfide content threshold, a Zn(Ac)2 solution with a preset volume ratio is injected to generate ZnS precipitate;

[0100] The dual-laser SERS technology enables full-spectrum detection within a preset frequency band, covering the blind spots of traditional methods.

[0101] Specifically, multimodal excitation detection is performed based on dual-laser SERS technology. A 785nm laser with a power of 50mW and a linewidth of less than 0.1nm is used to excite microplastics larger than 10μm through Rayleigh scattering. A 532nm laser with a power of 30mW and a linewidth of less than 0.1nm is used to excite microplastics of 1-10μm through surface-enhanced Raman effect.

[0102] Turbidity of microplastic samples was obtained using online sensors. The original sensor signal intensity is corrected based on the turbidity of the microplastic sample, using the correction formula:

[0103]

[0104] The corrected signal strength is calculated. ,in, The original sensing signal strength, It is a natural constant. For the water temperature of the microplastic sample, Turbidity of microplastic samples The temperature effect coefficient is preset to 0.001, and the correction formula is obtained by calibration using 100 sets of standard microplastic samples under different turbidities.

[0105] Concentration inversion is performed by constructing a segmented model based on the strength of the corrected signal;

[0106] If the corrected signal strength is less than the signal strength threshold ,and Then, using the concentration inversion formula: The concentration of microplastics was calculated, where, The standard sample fit factor, This represents the real-time power of the laser. To calibrate the power, To correct signal strength;

[0107] If the corrected signal strength is greater than or equal to the signal strength threshold, then the concentration inversion formula is used: The concentration of microplastics was calculated, where, The standard sample fit factor, This represents the real-time power of the laser. To calibrate the power, To correct signal strength, This is a concentration compensation term;

[0108] Step 3: Acquire and fuse multi-source data, correct abnormal federated values ​​based on the segmented model using a loss function and perform global aggregation, perform quantum evidence storage based on the aggregation completion, and simultaneously establish a self-healing sealed chamber and a dynamic temperature-controlled sample library to achieve non-destructive preservation of sample activity;

[0109] In a specific embodiment, SERS spectra, hydrological parameters, and in-situ images are acquired, and the sub-indices are normalized and then weighted and fused. The weights are determined by the analytic hierarchy process.

[0110] Based on the obtained SERS spectrum, hydrological parameters and in-situ image analysis, sub-indices were obtained. The sub-indices include: normalized value of microplastic concentration, normalized comprehensive score of hydrological parameters and normalized feature matching degree of image pollution. The comprehensive index was calculated by weighted fusion formula, and each sub-index was normalized by min-max.

[0111] Based on a comprehensive index combined with a local model, outlier federated values ​​are corrected and globally aggregated using a loss function;

[0112] Specifically, based on the local model, the loss function is:

[0113]

[0114] Correcting abnormal federated values, among which, For model parameters, For sub-index weights, These are measured values. These are the model's predicted values. The residual is the measured value.

[0115] Quantum evidence storage is performed based on global aggregation.

[0116] Generate hash: ,in, To provide timing for BeiDou, These are the coordinates of the sampling point. A quantum fingerprint signed with the manufacturer's private key. This is the raw data block, containing spectral and hydrological information;

[0117] Transmission: The data is transmitted via blue-green laser communication, with a preset wavelength of 520nm and a preset power of 1W, and uploaded to the quantum satellite to achieve tamper-proof evidence storage.

[0118] Constructing a self-healing sealed chamber: A trypsin solution at 40°C and pH 7.5 is introduced and circulated through a peristaltic pump for one cycle to degrade bio-attached organic matter; a 20 kPa pressure pulse is applied to continuously remove pipeline residues, with the pressure provided by a micro plunger pump; and the ZrO2 coating is irradiated with UV-A to utilize hydroxyl radicals to degrade residual organic matter.

[0119] Constructing a dynamic temperature-controlled sample library: Utilizing a temperature control algorithm, through the formula:

[0120]

[0121] The calculated sample library control temperature ,in, The optimal storage temperature for contaminants, Correction factor for container heat capacity The material of the cabin determines the composition. The default value is 0.8. For ambient temperature, The temperature coefficient of degradation rate, For estimated storage time;

[0122] Phase change material chamber maintenance ±0.3℃, within a preset recording period, the temperature data of the sample library is periodically recorded through blockchain to form an immutable temperature traceability chain;

[0123] Example 2: As Figure 2 As shown in the figure, the sampling system for water environment monitoring provided in this embodiment of the invention specifically includes the following modules:

[0124] Pre-inspection planning module: Introduces a digital twin pre-inspection system and a quantum encrypted blockchain calibration chain to pre-certify the trustworthiness of sensors. Pre-inspection is carried out through a bionic sampling head to obtain the historical error standard deviation of sensors, calculate the parameter decision weight to quantify sensor trustworthiness, obtain the target point coordinates to establish a flight time model, calculate the flight time of the sampling device and the sampling new course to plan a dynamic path for sampling;

[0125] Layered fixing module: Based on the planned dynamic path, it uses counter-current shearing layering technology for laminar flow control sampling, calculates the critical settling velocity to achieve zero mixing between layers, performs on-chip fixing decision based on laminar flow control sampling, and constructs a segmented model by calculating the correction signal strength to invert the concentration value of the detected object.

[0126] The fusion and evidence storage module acquires and fuses multi-source data, corrects abnormal federated values ​​based on a segmented model using a loss function, and performs global aggregation. Quantum evidence storage is then performed based on the aggregation, and a self-healing sealed chamber and a dynamic temperature-controlled sample library are established simultaneously to achieve non-destructive preservation of sample activity.

[0127] Example 3: A sampling device for water environment monitoring provided in this embodiment of the invention specifically includes the following:

[0128] This device includes: a digital twin module, a quantum encrypted blockchain module, and a dual-laser SERS detection unit; the sampling head adopts a biomimetic gill structure with three symmetrically distributed gill lobes, a single-lobe opening area, and built-in fluorescent microsphere detection channels and self-cleaning components; the intelligent control system integrates a federated learning framework and a four-dimensional hydrodynamic algorithm; the sample storage chamber is upgraded to a self-healing sealed structure and equipped with a phase change material temperature control system; the power system uses a vector thruster, supports heading adjustment, and meets the requirements of countercurrent stratified sampling.

[0129] The above provides a detailed description of one embodiment of the present invention, but the content described is only a preferred embodiment of the present invention and should not be considered as limiting the scope of the present invention. The above formulas are all dimensionless numerical calculations, and the formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art based on actual conditions and historical experience, and can be adjusted according to actual conditions. The above descriptions are only preferred embodiments of the present invention and are not intended to limit the present invention. All equivalent changes and improvements made in accordance with the scope of the present invention should still fall within the patent coverage of the present invention.

Claims

1. A sampling method for detecting a water environment, characterized by, The method comprises the following steps: The digital twin pre-inspection system and the quantum encryption blockchain calibration chain are introduced to pre-authenticate the sensor credibility, pre-inspection is performed through the bionic sampling head, the historical error standard deviation of the sensor is obtained, the parameter decision weight is calculated to quantify the sensor credibility, the target point coordinates are obtained to establish a navigation time model, the navigation time of the sampling device is calculated, a new sampling direction is planned, a dynamic path is planned, and sampling is performed; Based on the planned dynamic path, the laminar flow control sampling is performed by using the counter-current shear layering technology, the critical settling flow rate is calculated to realize zero mixing between layers, on-line fixed decision is made based on the laminar flow control sampling, the concentration inversion is calculated based on the corrected signal strength to construct a segmented model, and the concentration value of the detected object is calculated; Multi-source data is obtained and fused, the loss function is used to correct the abnormal federal value based on the segmented model, and global aggregation is performed, quantum evidence is recorded based on the aggregation, a self-healing sealed cabin and a dynamic temperature control sample library are simultaneously established, and sample activity is preserved without damage; The method for calculating the concentration inversion is: Obtaining turbidity of microplastic sample by online sensor correcting original sensing signal intensity based on turbidity of microplastic sample, by correction formula: The calculated correction signal intensity wherein, is the original sensing signal intensity, is a natural constant, is the water temperature of the microplastic sample, is the turbidity of the microplastic sample, is the temperature influence coefficient; A segmented model is constructed based on the corrected signal strength to calculate the concentration inversion; The method for correcting the abnormal federal value and performing global aggregation is: SERS spectrum, hydrological parameters, and in-situ images are obtained, sub-indices are normalized and analyzed to obtain sub-indices, and a comprehensive index is calculated through a weighted fusion formula; The loss function is used to correct the abnormal federal value based on the comprehensive index and the local model, and global aggregation is performed; The loss function is used based on the local model: correcting for abnormal federal values, wherein, is a model parameter, is a sub-index weight, is a measured value, is a model predicted value, is a measured value residual.

2. The sampling method for detecting a water environment according to claim 1, characterized by, The method for pre-inspection is: The bionic sampling head is used for pre-inspection, the sampling head adopts a bionic gill structure, a test liquid containing fluorescent microspheres is introduced into the sampling head, the microspheres are made of polystyrene, and the passing rate of the microspheres is counted in real time; Computing flow rate wherein is the theoretical microsphere throughput, is the test fluid flow rate, is the total gill flap opening area, is the actual microsphere throughput; If Greater than 95%, the system automatically starts ultrasonic cleaning, repeated detection to reach the standard; Monthly averages of pollution data in the sampling area in the past three years are obtained, the sensor calibration certificate is called by the blockchain, and the SHA-256 hash algorithm is used for verification, and the hash value is compared with the database of the manufacturer. If the hash values are consistent, the verification is passed.

3. The sampling method for detecting a water environment according to claim 1, characterized by, The method for calculating the parameter decision weight is: The historical error standard deviation of the sensor is obtained, and the formula is: The parameter decision weight is calculated wherein, is the sensor historical error standard deviation, is the sensor annual drift rate; if the sensor historical error standard deviation is greater than 0.1, it is automatically marked as a backup state.

4. The sampling method for detecting a water environment according to claim 1, characterized by, The method for calculating the navigation time and the new sampling direction is: A four-dimensional hydrodynamic positioning algorithm is used to fuse the flow velocity vector and turbulence intensity to obtain the target point coordinates , and a dynamic path is planned for sampling. Acquiring and inputting target point coordinates , acquiring three-dimensional flow field in real time through ADCP flow meter ; The navigation time model is established based on the target point coordinates, and the formula is: Computed sailing time wherein, is the straight-line distance to the target point and , is the average flow velocity vector module, is the turbulence intensity, calculated by the flow velocity standard deviation, is the angle between the flow direction and the sailing path, is the medium turbulence coefficient, is the maximum acceleration of the propeller; Based on the calculated navigation time, the underwater digital elevation model (DEM) is generated by using the multi-beam sonar to adapt to the terrain and avoid obstacles; If the distance d is less than the preset safety distance threshold, the three-dimensional coordinates of the surface points of the obstacle are determined by the multi-beam sonar, and the unit vector of the surface normal is obtained by geometric fitting to extract the obstacle normal vector Based on the obstacle normal vector, the formula: The new heading is calculated wherein, is the original heading, is the obstacle roughness coefficient, is the obstacle normal vector.

5. The sampling method for detecting a water environment according to claim 1, characterized by, The method for calculating the critical settling flow rate is: The temperature sensor array data is used for laminar flow control sampling, the target depth is determined by combining the thermocline algorithm, and the formula is: The critical settling velocity is calculated as wherein, g is the acceleration of gravity, ρp is the density of the particles, ρw is the density of the water, dp is the diameter of the particles, C is the drag coefficient; Based on the obtained critical settling flow rate, the sampling head is adjusted by the vector thruster at an angle of 15° to cut in, the gill opening area is dynamically adjusted, and the water entry speed is stabilized in the preset range through real-time feedback adjustment by the Doppler flowmeter.

6. The sampling method for detecting a water environment according to claim 1, characterized by, The method for making on-line fixed decision is: For heavy metal detection: when the ORP sensor reading is less than the preset reading threshold, the microfluidic chip automatically injects 65% purity AR grade HNO3, the injection amount is adjusted through the pH sensor feedback, and the target pH value is less than 2; For volatile organic compounds: when the temperature sensor detects that the temperature is greater than the preset temperature threshold, the Peltier element starts refrigeration, and the sample chamber is lowered to the preset refrigeration temperature within the preset refrigeration period and triggers the electromagnetic sealing valve to close; For sulfides: when the DO sensor reading is less than the sulfide content threshold, a preset volume ratio of Zn(Ac)2 solution is injected to generate ZnS precipitate; Full-spectrum detection in the preset frequency band is realized through double-laser SERS technology.

7. The sampling method for detecting a water environment according to claim 1, characterized by, The method of the dynamic temperature control sample library is: Constructing a dynamic temperature control sample library: using a temperature control algorithm, through the formula: The sample library control temperature is calculated wherein, is the optimal preservation temperature for the pollutant, is the heat capacity correction factor for the container, further , is the ambient temperature, is the temperature coefficient of the degradation rate, is the expected preservation time; The phase change material cabin maintains a preset range, and the temperature data of the sample library is recorded periodically through the block chain within a preset recording period, forming an unalterable temperature traceability chain.

Images