A multi-parameter carbonate system determination system and method

Abstract

The application discloses a kind of multi-parameter carbonate system determination system and method.The liquid path unit of system transports sample mixed system to be determined to the reaction cell based on conical guide;Gas path unit includes condensation purification module based on dry optimization and switchable carrier gas module, sample mixed system and the gas mixed after carrier gas module output obtains reaction product, and the extraction purification of reaction product is completed again through condensation purification module;Control unit receives the initial data of carbonate system collected by sensor detection unit, and carries out time window matching and fusion processing to carry out carbonate system parameter determination.The application can realize the sample processing of liquid path and gas path cooperation, sensor detection of multi-parameter cooperative acquisition, automatic calculation and control of software and hardware core, multi-parameter coupling and multi-time scale data processing, and dual-mode design for multi-scene application, can adapt to different carrier gas conditions and sample supply mode, so as to expand system application scene.

Description

Technical Field

[0001] This invention relates to a carbonate system determination system, and to the field of water quality monitoring and carbonate system analysis technology, specifically to a multi-parameter carbonate system determination system and method. Background Technology

[0002] In environmental monitoring and aquatic carbon cycle research, accurate determination of carbonate system parameters is crucial for analyzing aquatic carbon balance and its response to environmental changes. Typically, a complete characterization of the carbonate system requires measuring parameters such as dissolved inorganic carbon (DIC), pH, temperature, salinity, total alkalinity, and partial pressure of carbon dioxide, and then calculating some derived parameters.

[0003] In existing technologies, the above parameters are typically measured step-by-step by multiple independent devices, and then processed and calculated offline by manual labor or computers. This method is not only complex and lacks automation, but it is also prone to human error, making it difficult to meet the needs of continuous monitoring and field applications.

[0004] In particular, dissolved inorganic carbon (DIC) is generally obtained by integrating the CO2 concentration released after sample acidification over time; while pH, temperature, and salinity are usually collected continuously at a higher sampling frequency, reflecting instantaneous states. In existing technologies, a specific moment value or simple statistical value is often directly selected and combined with the DIC result obtained in that period into the carbonate system calculation model. This leads to inconsistencies in the input parameters in terms of time reference and physical meaning, and makes them susceptible to environmental abrupt changes and sensor instability, thus affecting the accuracy and stability of the calculated results for total alkalinity, carbon dioxide partial pressure, and other carbonate system parameters.

[0005] In addition, existing carbonate system measuring devices are mostly designed for laboratory environments, and lack the ability to perform real-time on-site measurements and switch between multiple scenarios.

[0006] Therefore, it is necessary to provide a multi-parameter carbonate system determination system and method to achieve the coordinated acquisition of dissolved inorganic carbon (DIC) and high-frequency auxiliary parameters, and to perform unified time reference processing and consistency mapping on asynchronous sampling data, thereby improving the automation, accuracy and stability of carbonate system parameter determination. Summary of the Invention

[0007] To address the problems existing in the background art, this invention provides a multi-parameter carbonate system determination system and method. This invention solves the problems of existing technologies where carbonate system parameters rely on multi-device step-by-step measurement, offline calculation, and inconsistent time bases for different parameters. It enables multi-parameter collaborative measurement, consistent processing of asynchronous sampling data, and automatic calculation of carbonate system parameters, making it suitable for laboratory and field monitoring scenarios. The system integrates measurements of parameters such as dissolved inorganic carbon (DIC), pH, temperature, and salinity, and performs unified time base processing, window mapping, and fusion calculation on data acquired at different sampling frequencies to achieve automatic determination of carbonate system parameters.

[0008] The technical solution adopted in this invention is: I. A multi-parameter carbonate system determination system, comprising: The liquid circuit unit is used to precisely deliver microliter-level sample mixtures to the reaction cell based on a conical flow path while simultaneously recovering the liquid.

[0009] The gas path unit includes a drying-optimized condensation and purification module and a switchable carrier gas module. The carrier gas module and the liquid path unit are connected to the bottom of the reaction cell. The sample mixing system and the gas output from the carrier gas module are mixed to obtain the reaction product. The condensation and purification module is connected to the top of the reaction cell and is used to jointly complete the extraction and purification of the reaction product and recover the gas at the same time.

[0010] The sensor detection unit is installed in the liquid circuit unit and the condensation and purification module to collect initial data of the carbonate system.

[0011] The control unit, electrically connected to the liquid circuit unit, carrier gas module, and condensation and purification module, is used to receive the initial data of the carbonate system, perform time window matching and recursive fusion processing, and then obtain the parameters of the carbonate system to realize the determination of multi-parameter carbonate system.

[0012] The sensor detection unit includes an infrared detector, a pH sensor, and a temperature and salinity probe. The infrared detector is installed in the condensation and purification module to collect data on the CO2 concentration change of the flowing gas during the measurement period, and then inputs the data into the control unit to obtain the dissolved inorganic carbon (DIC) result through integration. The pH sensor and temperature and salinity probe are installed in the liquid circuit unit to collect the pH value, temperature, and salinity data of the flowing water body during the measurement period. The sampling frequency of the pH sensor and temperature and salinity probe is higher than the data update frequency of the dissolved inorganic carbon (DIC) result. The infrared detector, pH sensor, and temperature and salinity probe are all electrically connected to the control unit.

[0013] The reaction tank has a transparent acrylic cylinder in the middle. Both the top and bottom of the tank have coaxial conical guide ports that run through the middle, connecting the condensation and purification module and the liquid circuit unit respectively. The diameter of the two conical guide ports closer to the middle of the reaction tank is larger than that of the side farther away from the middle. A through hole is opened on the side of the bottom of the tank to connect the gas circuit unit.

[0014] The liquid circuit unit includes an injection pump, a waste liquid tank, a water pump, and a sample cell. Measurements are performed in two modes: laboratory mode and mobile mode. In laboratory mode, the sample cell directly stores the water sample to be measured. In mobile mode, the water pump extracts the water sample to be measured from the mobile environment, and the water sample in the sample cell is directly discharged back into the mobile environment. A pH sensor and a temperature and salinity probe are installed in the sample cell to measure the pH, temperature, and salinity data of the water sample. The various inlets of the injection pump's liquid inlet area are connected via pipes to the sample cell and a container storing phosphoric acid solution and distilled water. The sample mixing system consists of the water sample to be measured, phosphoric acid solution, and distilled water. Another inlet of the injection pump is connected to the conical guide port at the bottom of the reaction tank, and the output port of the injection pump is connected to the waste liquid tank to recover the liquid from the injection pump. Both the injection pump and the water pump are electrically connected to the control unit.

[0015] The carrier gas module includes a carrier gas device, a second drying tube, and a flow meter. In laboratory mode, the carrier gas device uses a nitrogen cylinder as the gas source. The nitrogen cylinder, the second drying tube, and the through hole at the bottom of the reaction tank are connected in sequence. The flow meter is installed on the pipe between the second drying tube and the bottom of the reaction tank. In mobile mode, the carrier gas device uses an air pump as the gas source. The air pump is directly connected to the bottom of the reaction tank. The flow meter is installed on the pipe between the air pump and the bottom of the reaction tank. The flow meter detects the flowing gas to control the gas to flow at a preset speed. A nitrogen cylinder is used in laboratory mode for high precision, while an air pump is used in mobile mode for portability. The sample mixing system and the nitrogen or air output from the carrier gas module are mixed to obtain the reaction product. Both the air pump and the flow meter are electrically connected to the control unit.

[0016] The condensation purification module includes a condenser, a first drying tube, a filter, a three-way valve, and a waste gas cylinder connected in sequence. The first outlet of the three-way valve is directly connected to the waste gas cylinder, and the second outlet of the three-way valve is connected to the waste gas cylinder via an infrared detector. The condenser is electrically connected to a power supply, and the three-way valve is electrically connected to a control unit.

[0017] The interior of the first drying tube is a conical flow guide space that is narrow at both ends and wide in the middle, in order to improve dehumidification efficiency and airflow stability.

[0018] II. A determination method for a multi-parameter carbonate system, comprising: Step 1) During the measurement period, the water sample to be measured is extracted multiple times. The control unit receives the initial data of the carbonate system, including the CO2 concentration change data of the reaction products during the sample acidification and purging process, as well as the pH value, temperature and salinity data of the water sample to be measured. The pH value, temperature and salinity data of the water sample to be measured are all used as the initial measurement data. After recursively fusion processing of each initial measurement data, the preprocessed measurement data is obtained. At the same time, the dissolved inorganic carbon (DIC) result is obtained based on the CO2 concentration change data.

[0019] Step 2) Based on the dissolved inorganic carbon (DIC) results, several reaction time windows are determined. For each pre-processed measurement data and each reaction time window, the data of the pre-processed measurement data within the reaction time window are extracted and mapped to the reaction time window. Then, after weighted statistical processing, the parameter characterization values ​​that characterize the water state of the water sample to be measured are obtained. The dissolved inorganic carbon (DIC) results and parameter characterization values ​​within each reaction time window are constructed into a multi-parameter synchronous data set.

[0020] Step 3) Input the multi-parameter synchronous data set into the carbonate system calculation model, process it, and output the carbonate system parameters.

[0021] In step 1), the recursive fusion processing specifically employs an estimation method based on state prediction and observation update to iteratively process the initial measurement data. Each iteration outputs a predicted value, and the residual between the predicted value and the initial measurement data is used as the dynamic residual of the estimation method based on state prediction and observation update. The process noise covariance and measurement noise covariance are dynamically adjusted based on the dynamic residual to adaptively correct the initial measurement data, thus obtaining preprocessed measurement data. The dynamic residual can be used for anomaly detection; if the dynamic residual exceeds a preset threshold, it is determined that the sensor has an anomaly or is drifting.

[0022] In step 2), the starting point of the reaction time window is the moment when the CO2 concentration in the dissolved inorganic carbon (DIC) result exceeds the baseline range, and the ending point is the moment when the CO2 concentration returns to the baseline range. When the gas path unit is not connected to the liquid path unit, the baseline range of CO2 concentration is obtained by detecting the infrared detector. During weighted statistical processing, outliers must first be removed. Then, for each pre-processed measurement data within each reaction time window, the weighted mean of the pre-processed measurement data is obtained. The weight of the data from the start time to the end time of the reaction time window in the pre-processed measurement data gradually decreases. Finally, the weighted mean of each pre-processed measurement data in each reaction time window is averaged and used as the parameter characterization value.

[0023] The key technical problem this invention aims to solve is not merely to achieve automatic measurement of parameters such as dissolved inorganic carbon (DIC), pH, temperature, and salinity, but rather that: DIC is obtained by integrating a CO2 concentration curve, and its result corresponds to the comprehensive response within a measurement period; while pH, temperature, and salinity are typically collected continuously at a high sampling frequency, reflecting instantaneous states, and the two are not naturally consistent in terms of time scale and physical meaning. Therefore, this invention constructs a multi-parameter carbonate system measurement platform integrating a liquid path unit, a gas path unit, a sensor detection unit, and a control unit 11. Using the DIC measurement period as a unified time reference, it performs window observation mapping and fusion processing on high-frequency parameter data within the measurement period to generate parameter characterization values ​​corresponding to the DIC results, thereby constructing a multi-parameter synchronous data set corresponding to the same system state.

[0024] Furthermore, under a unified time reference, this invention employs an estimation method based on state prediction and observation updates to asynchronously update, detect anomalies, and adaptively correct data at different sampling frequencies to obtain a multi-parameter synchronous data set characterizing the state of the same system. This multi-parameter synchronous data set is then input into a carbonate system calculation model to calculate carbonate system parameters such as total alkalinity and partial pressure of carbon dioxide.

[0025] The beneficial effects of this invention are: 1. Sample processing using a combination of liquid and gas pathways: The liquid and gas circuit units work together to provide optimized conditions for acidification and gas extraction. The liquid circuit unit, through precise multi-channel control of the syringe pump, enables automatic addition of samples and reagents (such as 8.5% phosphoric acid and standards) and automatic cleaning of the reaction chamber. The gas circuit unit features a switchable carrier gas source and includes purification, flow control, and dehumidification functions. The liquid and gas circuit units work together within the reaction chamber to complete the acidification and gas extraction. The reaction chamber employs a three-part combined design, with its lower conical inlet and outlet integrated, which improves gas-liquid mixing efficiency and reduces dead volume, providing a structural basis for DIC measurement.

[0026] 2. Sensor detection with multi-parameter collaborative acquisition: The sensor detection unit, through collaborative acquisition and fusion processing of multi-parameter data, helps improve the consistency and comparability of basic measurement data. State estimation and anomaly correction enhance data stability and reliability. Furthermore, estimation methods based on state prediction and observation updates can be used to smooth multi-parameter data, identify anomalies, and compensate for drift.

[0027] 3. The core hardware and software components for achieving automatic calculation and control: The control unit includes a lower-level machine and upper-level software. The lower-level machine is used to execute hardware control, while the upper-level software is used to implement equipment control, data acquisition, data processing, and result display. By integrating the data processing module with the carbonate system calculation model, manual data integration and offline calculation steps can be reduced, thereby improving the unit's automation level and the stability of calculation results.

[0028] 4. Multi-parameter coupling and multi-timescale data processing: In terms of data processing, this invention uses the DIC measurement cycle as a unified time reference to perform window mapping and recursive fusion processing on the high-frequency collected pH, temperature and salinity data, and improves the stability of the construction of multi-parameter synchronous data sets under asynchronous sampling conditions through residual discrimination and noise covariance adaptive adjustment.

[0029] 5. Dual-mode design for multi-scenario applications: It supports both laboratory and mobile working modes and can adapt to different carrier gas conditions and sample supply methods, thereby expanding the applicable scenarios of the unit. Attached Figure Description

[0030] Figure 1 This is a schematic diagram illustrating the overall principle of the system of the present invention; Figure 2 This is a schematic diagram of the overall system structure; Figure 3 This is a schematic diagram of the reaction tank and condenser of the present invention, wherein, Figure 3 (a) is a schematic diagram of the reaction tank. Figure 3 (b) is a schematic diagram of the condenser structure; Figure 4 The images show a comparison of simulation results for the drying tubes. Figure 4 (a) is a schematic diagram of the airflow distribution before the internal structure of the drying tube was optimized. Figure 4 (b) is a schematic diagram of the airflow distribution after the internal structure of the drying tube has been optimized; Figure 5 A graph showing the results of five-point calibration for dissolved inorganic carbon (DIC) measurements; Figure 6 The figure shows the repeatability test results for dissolved inorganic carbon (DIC), where (a) is the measurement result with nitrogen as the carrier gas and (b) is the measurement result with air as the carrier gas. Figure 7 A comparison chart of calculation errors for carbonate system parameters after processing raw data using different methods; In the diagram: 1. Liquid inlet area, 2. Injection pump, 3. Reaction tank, 4. Condenser, 5. First drying tube, 6. Filter, 7. Three-way valve, 8. Carrier gas device, 9. Second drying tube, 10. Flow meter, 11. Control unit, 12. Infrared detector, 13. Waste liquid tank, 14. Waste gas cylinder, 15. pH sensor, 16. Temperature and salinity probe, 17. Water pump. Detailed Implementation

[0031] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0032] like Figure 1 and Figure 2 As shown, the parameter carbonate system determination system of the present invention includes a liquid path unit, a gas path unit, a sensor detection unit, and a control unit 11. The liquid path unit is used to accurately deliver the micro-level sample mixture to be measured to the reaction cell 3 based on a conical flow guide and simultaneously recover the liquid. The gas path unit includes a drying-optimized condensation and purification module and a switchable carrier gas module. The carrier gas module and the liquid path unit are connected to the bottom of the reaction cell 3. The sample mixture and the gas output from the carrier gas module are mixed to obtain the reaction product. The condensation and purification module is connected to the top of the reaction cell 3 and is used to jointly complete the extraction and purification of the reaction product and simultaneously recover the gas. The sensor detection unit is installed in the liquid path unit and the condensation and purification module to collect the initial data of the carbonate system. The control unit 11 is electrically connected to the liquid path unit, the carrier gas module, and the condensation and purification module. It is used to receive the initial data of the carbonate system, perform time window matching and recursive fusion processing, and then obtain the carbonate system parameters to realize the determination of multi-parameter carbonate system.

[0033] The multi-parameter carbonate system measurement system achieves coordinated acquisition of dissolved inorganic carbon (DIC), pH value, temperature, and salinity through the coordinated operation of the liquid circuit unit, gas circuit unit, sensor detection unit, and control unit 11. It also performs time window matching, fusion processing, and carbonate system parameter calculation on the data acquired at different sampling frequencies.

[0034] The sensor detection unit includes an infrared detector 12, a pH sensor 15, and a temperature and salinity probe 16. The infrared detector 12 is installed in the condensation and purification module to collect data on the CO2 concentration change of the flowing gas during the measurement period, and then inputs the data into the control unit to obtain the dissolved inorganic carbon (DIC) result through integration. The pH sensor 15 and the temperature and salinity probe 16 are installed in the liquid circuit unit to collect data on the pH value, temperature, and salinity of the flowing water during the measurement period. The sampling frequency of the pH sensor 15 and the temperature and salinity probe 16 is higher than the data update frequency of the dissolved inorganic carbon (DIC) result. The infrared detector 12, the pH sensor 15, and the temperature and salinity probe 16 are all electrically connected to the control unit 11.

[0035] The sensor detection system is responsible for acquiring multi-parameter data and forms the data foundation for analyzing carbonate systems. This system includes, but is not limited to: a non-dispersive infrared (NDIR) detector 12 for measuring the concentration of escaping CO2, a pH sensor 15 for measuring the pH value of the water, and a temperature-salinity probe 16 for measuring the temperature and salinity of the water sample. The pH sensor 15 and the temperature-salinity probe 16 continuously acquire water sample parameters at a high sampling frequency, while the NDIR detector 12 continuously acquires the CO2 concentration change curve during the sample acidification and purging process, and calculates the dissolved inorganic carbon (DIC) results for the corresponding measurement period through integration.

[0036] like Figure 3 As shown in (a), the middle part of the reaction tank 3 is a transparent acrylic cylinder. Both the top and bottom of the tank have coaxial conical inlets that penetrate the middle, connecting to the condensation and purification module and the liquid path unit, respectively. The diameter of the two conical inlets closer to the middle of the reaction tank 3 is larger than that of the inlets further away from the middle. A through-hole is opened on the side of the bottom of the tank, penetrating the middle, to connect to the gas path unit. Specifically, the reaction tank 3 adopts a three-part composite structure: a central transparent acrylic cylinder, upper and lower parts made of 3D-printed cyclohexanediol modified copolyester PETG (Polyethylene terephthalate glycol-modified). It also features conical inlets. This structure combines the inlet and outlet liquid outlets, which helps improve the gas-liquid mixing effect within the reaction tank 3 and reduces the residual dead volume of the liquid, thereby improving CO2 release and extraction efficiency and providing a structural basis for DIC measurement.

[0037] The liquid circuit unit includes a syringe pump 2, a waste liquid tank 13, a water pump 17, and a sample cell. Measurements are performed in laboratory mode and mobile mode. In laboratory mode, the sample cell directly stores the water sample to be measured. In mobile mode, the water pump 17 extracts the water sample to be measured from the mobile environment, and the water sample in the sample cell is directly discharged back into the mobile environment. A pH sensor 15 and a temperature and salinity probe 16 are installed in the sample cell to measure the pH, temperature, and salinity data of the water sample. The liquid inlet area 1 of the syringe pump 2 has several transmission ports connected to the sample cell and containers storing phosphoric acid solution and distilled water, respectively, via pipes. The phosphoric acid solution has a mass fraction of 8.5%. The liquid inlet area 1 also includes several input ports to allow different liquids to be introduced in different measurement experiments. The sample mixing system consists of the water sample to be measured, phosphoric acid solution, and distilled water. Another transmission port of the syringe pump 2 is connected to the conical guide port at the bottom of the reaction cell 3. The output port of the syringe pump 2 is connected to the waste liquid tank 13 to recover the liquid in the syringe pump 2. Both the syringe pump 2 and the water pump 17 are electrically connected to the control unit 11. The core function of syringe pump 2 is to achieve precise delivery and switching of liquids at micro-liter levels. Specifically, syringe pump 2 first draws distilled water to clean reaction tank 3, and then draws a water sample and a measured amount of phosphoric acid to inject into reaction tank 3 for acidification reaction.

[0038] The carrier gas module includes a carrier gas device 8, a second drying tube 9, and a flow meter 10. In laboratory mode, the carrier gas device 8 uses a nitrogen cylinder as the gas source. The nitrogen cylinder, the second drying tube 9, and the through hole at the bottom of the reaction tank 3 are connected in sequence. The flow meter 10 is installed on the pipe between the second drying tube 9 and the bottom of the reaction tank 3. The second drying tube 9 is filled with a mixture of soda lime and calcium chloride as a desiccant. In mobile mode, the carrier gas device 8 uses an air pump as the gas source. The air pump is directly connected to the bottom of the reaction tank 3. The flow meter 10 is installed between the air pump and the bottom of the reaction tank 3. On the pipeline, the flow meter 10 detects the gas flowing through it to control the gas to flow at a preset speed. In laboratory mode, a nitrogen cylinder is used to pursue high precision, while in mobile mode, an air pump is used for convenience and portability. The sample mixing system and the nitrogen or air output from the carrier gas module are mixed to obtain the reaction product. The carrier gas of the carrier gas module is purified by the second drying tube 9 and then controlled by the flow meter 10 to be introduced into the reaction cell 3 at a constant flow rate, such as 100 mL / min, to carry out the CO2 gas generated by the acidification reaction. The air pump and the flow meter 10 are both electrically connected to the control unit 11.

[0039] The condensation and purification module includes a condenser 4, a first drying tube 5, a filter 6, a three-way valve 7, and a waste gas cylinder 14 connected in sequence. The first outlet of the three-way valve 7 is directly connected to the waste gas cylinder 14, and the second outlet of the three-way valve 7 is connected to the waste gas cylinder 14 via an infrared detector 12. The condenser 4 is electrically connected to a power source, and the three-way valve 7 is electrically connected to a control unit 11. The exhaust gas from the reaction tank 3 then flows through the condenser 4, which is cooled to 2°C by a semiconductor refrigeration chip, and the first drying tube 5, which contains calcium chloride, to remove moisture. Figure 3 As shown in (b), it is a semiconductor refrigeration chip.

[0040] The interior of the first drying tube 5 is a tapered airflow space that is narrow at both ends and wide in the middle to improve dehumidification efficiency and airflow stability; the second drying tube 9 has the same structure. Figure 4 (a) and Figure 4 As shown in (b), the conical space is optimized through computational fluid dynamics (CFD) simulation. The air path of the ordinary cylindrical drying tube is easily blocked, resulting in dead volume. The air path of the conical guide tube is smoother, which is conducive to improving dehumidification efficiency and airflow stability, and reducing the impact of water vapor on the detection process.

[0041] The determination method of the multi-parameter carbonate system determination system of the present invention is as follows: Step 1) During the measurement period, the water sample to be measured is drawn multiple times. The control unit 11 receives the initial data of the carbonate system, including the CO2 concentration change data of the reaction products during the sample acidification and purging process, as well as the pH value, temperature and salinity data of the water sample to be measured. The pH value, temperature and salinity data of the water sample to be measured are all used as the initial measurement data. After recursively fusion processing of each initial measurement data, the preprocessed measurement data is obtained. At the same time, the dissolved inorganic carbon (DIC) result is obtained based on the CO2 concentration change data. The measurement period corresponding to the DIC result is used as a unified time reference. The high-frequency collected pH, temperature and salinity data are used to continuously update the state estimation. The DIC result is used to periodically correct the state.

[0042] The recursive fusion processing specifically employs an estimation method based on state prediction and observation updates. This involves using a Kalman filter with dynamic residuals to iteratively process the initial measurement data. Each iteration outputs a predicted value, and the residual between the predicted value and the initial measurement data is used as the dynamic residual of the estimation method. Based on this dynamic residual, the process noise covariance and measurement noise covariance are dynamically adjusted to adaptively correct the initial measurement data, resulting in preprocessed measurement data. The dynamic residuals can also be used for anomaly detection; if the dynamic residuals exceed a preset threshold, it indicates that the sensor is malfunctioning or drifting.

[0043] Step 2) Based on the dissolved inorganic carbon (DIC) results, several reaction time windows are determined. For each pre-processed measurement data and each reaction time window, the data of the pre-processed measurement data within the reaction time window are extracted and mapped to the reaction time window. Then, after weighted statistical processing, the parameter characterization values ​​that characterize the water state of the water sample to be measured are obtained. The dissolved inorganic carbon (DIC) results and parameter characterization values ​​within each reaction time window are constructed into a multi-parameter synchronous data set.

[0044] The starting point of the reaction time window is the moment when the CO2 concentration in the dissolved inorganic carbon (DIC) results exceeds the baseline range, and the ending point is the moment when the CO2 concentration returns to the baseline range. When the gas path unit is not connected to the liquid path unit, the baseline range of CO2 concentration is obtained by infrared detector 12. During weighted statistical processing, outliers must first be removed. Then, for each pre-processed measurement data within each reaction time window, the weighted mean of the pre-processed measurement data is obtained. The weight of the data from the start time to the end time of the reaction time window in the pre-processed measurement data gradually decreases. Finally, the weighted mean of each pre-processed measurement data in each reaction time window is averaged and used as the parameter characterization value.

[0045] The control unit 11 is used to automatically control the working status of the liquid circuit unit and the gas circuit unit according to the preset measurement process, including cleaning, sample injection, acidification reaction, gas purging and measurement steps, and to switch processes based on the measurement status or data status. The control unit 11 adaptively adjusts the measurement process based on the stability of sensor data or abnormal detection results, including repeating measurements, process rollback or recalibration.

[0046] Step 3) Input the multi-parameter synchronous data set into the carbonate system calculation model, process it, and output the carbonate system parameters, including total alkalinity (TA) and partial pressure of carbon dioxide (pCO2).

[0047] The control unit 11 is used for unified scheduling and time-series management of multi-sensor data, establishing a correspondence between the dissolved inorganic carbon (DIC) measurement results and the pH, temperature, and salinity data within the same measurement period. Before entering the control unit 11, the raw multi-parameter data preferably undergoes estimation processing based on state prediction and observation updates, preferably Kalman filtering, to achieve data smoothing, noise suppression, and anomaly drift identification, thereby improving the accuracy and stability of subsequent carbonate system parameter calculations.

[0048] The control unit 11 uses an ESP32-S3 microcontroller as its lower-level core, and controls actuators such as the injection pump 2, three-way valve 7, and flow meter 10 via serial communication, while also receiving data from all sensors. The control unit 11 includes upper-level software developed in Python for equipment control, data acquisition, data processing, and result display. The software provides a graphical user interface (GUI) for equipment control and real-time monitoring, and integrates a carbonate system calculation model. Using DIC measurement results and corresponding pH, temperature, and salinity data processed within a time window during the measurement period as input, it calculates a series of carbonate system parameters in real time, such as total alkalinity (TA) and carbon dioxide partial pressure (pCO2), and displays them directly on the GUI. This reduces manual data integration and offline calculation steps, achieving integrated measurement, processing, and result output.

[0049] In this embodiment, the host computer software, in addition to its functions of device control and result display, also integrates a multi-parameter coupled data processing module. This data processing module preferably employs a Kalman filter algorithm to process the raw data of pH, temperature, salinity, and CO2 concentration. Specifically, it constructs a state vector for continuously acquired multi-parameter measurements and predicts the current data based on historical states. When new sensor measurements are obtained, a corrected state estimate is obtained through measurement updates. Simultaneously, the residual between the predicted and measured values ​​is used to determine if there is abnormal sensor drift. When the residual exceeds a preset threshold, the process noise covariance and measurement noise covariance are dynamically adjusted to achieve adaptive correction of the measurement data.

[0050] In a preferred embodiment, the multi-parameter coupled data processing module can be implemented as follows: Within a DIC measurement cycle, the DIC results are obtained by integrating the CO2 concentration versus time curve, while pH, temperature, and salinity are continuously collected at a higher sampling frequency. To ensure that the high-frequency auxiliary parameters correspond to the system state within the same measurement cycle as the DIC results, the data processing module uses the DIC measurement cycle as a unified time reference, maps and statistically processes the high-frequency parameters within the time window, and generates parameter characterization values ​​corresponding to the DIC results.

[0051] To address the asynchronous sampling characteristics, the data processing module constructs a unified state space encompassing DIC, pH, temperature, and salinity, and performs recursive fusion processing on data from different time scales. Specifically, it defines the system state vector. for: in, These represent the dissolved inorganic carbon concentration, pH value, temperature, and salinity at time [time]. k The state.

[0052] 1) In the state prediction phase, the system uses the state estimate from the previous moment. Regarding the current state Make a prediction: in, F This is the state transition matrix, used to describe the changing trends of each parameter over time and the sensor drift characteristics; w k This is process noise, used to characterize environmental disturbances and system uncertainties.

[0053] 2) During the data update phase, construct the corresponding observation vector and observation matrix based on the measurement data available at the current moment to achieve asynchronous updates: When only high-frequency parameters (pH, temperature, salinity) are acquired, a partial observation vector is constructed. : Where T is the transpose.

[0054] Corresponding observation matrix H fast It is used to continuously update the state.

[0055] When acquiring DIC measurement results, construct the complete observation vector. : Corresponding observation matrix H full The system state is periodically corrected.

[0056] The system switches the observation matrix according to the observation type and updates the status accordingly. : in, For the current moment Available observation vectors, when only high-frequency parameters such as pH, temperature, and salinity are available. Pick When obtaining DIC measurement results, Pick ; For the observation matrix corresponding to the current observation vector, when Pick hour, Take H fast ,when Pick hour, Take H full ;K k This is the Kalman gain.

[0057] 3) Since DIC is an integral measurement, its physical meaning corresponds to the overall response of the system within a time window Δt. Therefore, this invention maps high-frequency parameters to this time window: Among them, DIC k For the dissolved inorganic carbon measurement result corresponding to the kth measurement cycle, t k t represents the end time of the k-th measurement cycle or the end of the integration window. k -Δt is the start time of the measurement cycle or the start point of the integration window, and Δt is the length of the time window corresponding to the DIC measurement cycle; CO2(t) is the CO2 concentration collected by the infrared detector at time t.

[0058] Simultaneously, equivalent statistical processing is performed on the high-frequency data within the window: in, , and These represent the equivalent statistical values ​​of pH, temperature, and salinity at time k, respectively. , and The values ​​are pH, temperature, and salinity at time t, respectively.

[0059] This leads to the construction of a unified observation vector: By using time window mapping and statistical processing, the obtained parameter characterization values ​​and DIC results are made to correspond to the system state within the same measurement period. This time window mapping process essentially constructs a window observation model, enabling the fusion of integral measurement results and instantaneous measurement parameters within a unified observation framework. Preferably, the statistical processing of high-frequency data within the time window can adopt a weighted statistical method, where the weight of observation data closer to the moment the sample enters the system or the start of the acidification reaction is higher than the weight of observation data closer to the end of the measurement period.

[0060] 4) Define residuals for: in, The predicted observation is obtained by mapping the current state estimate to the observation matrix, and is used to represent the observation result that the sensor should theoretically output under the current state estimate conditions.

[0061] When the following conditions are met: in, The residual covariance matrix is ​​used to characterize the uncertainty of the residuals between the predicted and actual observed values. This is a preset threshold. When the residual statistic exceeds the preset threshold... When this occurs, it is determined that the current observation data contains anomalies, disturbances, or sensor drift.

[0062] Furthermore, based on the residual versus noise covariance Make dynamic adjustments: in, This is a forgetting factor or smoothing coefficient used to adjust the weight of the cross product of historical measurement noise covariance and current residual in the measurement noise covariance update. .when As the value approaches 1, the measurement noise covariance update relies more on historical estimates; when... The smaller the value, the greater the impact of the current residual on the measurement noise covariance update.

[0063] After completing the time scale unification and data fusion, the parameter set is obtained. .

[0064] By adaptively adjusting the process noise covariance and / or measurement noise covariance, the stability of state estimation and the robustness of data processing under complex environmental conditions are improved.

[0065] Coupled with the calculation model of the carbonate system, the synchronization parameter set under a unified time base is finally obtained: Inputting this into a carbonate system calculation model yields parameters such as total alkalinity (TA) and carbon dioxide partial pressure (pCO2). in, The parameters of the carbonate system at time k; This is a calculation model for a carbonate system.

[0066] The above methods enable unified time base processing, anomaly correction, and fusion calculation of multi-source asynchronous data, thereby improving the stability of data processing and the reliability of analysis results under complex environmental conditions.

[0067] The measurement method of the present invention includes: acidifying the water sample and extracting CO2 to obtain DIC results; simultaneously continuously collecting pH, temperature and salinity data of the water sample; then using the DIC corresponding measurement period as a unified time reference, performing window matching, fusion processing and anomaly correction on high-frequency parameters; finally inputting the obtained multi-parameter synchronous data set into the carbonate system calculation model to obtain carbonate system parameters such as total alkalinity and partial pressure of carbon dioxide.

[0068] The specific measurement steps are as follows: S1: System Initialization: Place distilled water, the sample solution to be tested, and an 8.5% phosphoric acid solution into the corresponding containers on the sample stage. Turn on the host computer software interface and the carrier gas source (nitrogen or air). At this time, the system is powered on, and the pH sensor 15 and the temperature and salinity probe 16 begin to continuously monitor the pH, temperature, and salinity data of the water sample in the flow cell or reaction cell, preparing data for subsequent calculations.

[0069] S2: Pipeline purification: Start the control unit 11 via the host computer software, open the corresponding three-way valve 7, and let the carrier gas continuously flow into the system at a set flow rate (e.g., 100 mL / min) for 45 seconds. The purpose is to remove residual gas in the pipeline, ensure that the gas path is consistent with the carrier gas, and create an ideal environment for subsequent high-precision CO2 measurement.

[0070] S3: Cleaning and rinsing of reaction cell 3: Control unit 11 sends a command to control syringe pump 2 to first draw an appropriate amount of distilled water into reaction cell 3 and then discharge it to clean reaction cell 3 and avoid cross-contamination of samples. Subsequently, syringe pump 2 draws a small amount of the sample solution to be tested to rinse reaction cell 3 to ensure the representativeness of the samples measured later. All waste liquid is discharged into waste liquid tank 13.

[0071] S4: Acidification Reaction and Multi-Parameter Coordinated Acquisition: 1 mL of sample solution and 0.5 mL of phosphoric acid solution are precisely added to the reaction cell via syringe pump 2. The acidification reaction occurs immediately, and DIC in the sample is converted into CO2 gas. Simultaneously, carrier gas is used to purge the reaction solution at a constant flow rate, carrying away the CO2 gas.

[0072] During this measurement period, infrared detector 12 continuously detects the CO2 concentration in the gas, generating a CO2 concentration-time curve; pH sensor 15 and temperature-salinity probe 16 continuously collect pH, temperature, and salinity data of the water body at a higher sampling frequency. All sensor data are transmitted to the host computer software, but the parameters are not sampled at the same frequency; instead, they are collected in a coordinated manner under a unified measurement period to provide the raw data foundation for subsequent multi-timescale fusion. All sensor data are transmitted to the host computer software via serial port and are buffered, labeled, and processed according to the unified measurement period.

[0073] S5: Data processing and calculation of carbonate system parameters, specifically including: 1) Preprocess the raw data of pH, temperature, salinity and CO2 concentration. The preprocessing includes smoothing, anomaly detection and adaptive correction.

[0074] 2) Integrate the CO2 concentration-time curve obtained in step S4, and convert it according to the standard curve to obtain the DIC result corresponding to the measurement cycle.

[0075] 3) Using the measurement period corresponding to the DIC results as a unified time reference, a time window is constructed. The pH, temperature and salinity data collected continuously within the window are modeled for window observation to generate parameter characterization values ​​corresponding to the DIC results.

[0076] 4) Construct a multi-parameter synchronous data set corresponding to the same system state based on the DIC results and parameter characterization values, and input it into the carbonate system calculation model to obtain carbonate system parameters such as total alkalinity and carbon dioxide partial pressure.

[0077] 5) Output and store the raw measurement data, preprocessing results, and carbonate system parameter calculation results.

[0078] S6: Cyclic Measurement: Repeat steps S3 to S5 to achieve continuous automated measurement of multiple samples. The carbonate system parameters of all samples are automatically generated according to the above process, which can improve the efficiency of continuous measurement and reduce human processing and calculation errors.

[0079] The performance verification is as follows: In one embodiment, such as Figure 6 (a) and Figure 6 As shown in (b), the repeatability test results indicate that the relative standard deviation of DIC measurement is approximately 0.2% when nitrogen is used as the carrier gas and approximately 0.26% when air is used as the carrier gas; Figure 5 As shown, the linear correlation coefficient R of the standard curve 2The value is greater than 0.999. The above results demonstrate that the system of this invention has good measurement linearity and repeatability. Furthermore, the host computer software can process the acquired sensor data and call the carbonate system calculation model to perform parameter calculations; the results can be displayed and saved in the graphical user interface.

[0080] In one embodiment, the window mapping and asynchronous fusion processing method based on a unified time reference of the present invention was compared and verified. This method was compared with the single-value method of directly selecting high-frequency parameter values ​​at the beginning and end of the measurement period, and the averaging method of simply averaging high-frequency parameters within the window. The calculation errors of total alkalinity, carbon dioxide partial pressure, and other carbonate system parameters were evaluated. Figure 7 As shown, compared with other processing methods, the method of the present invention exhibits lower RMSE in the calculation of total alkalinity (TA) and carbon dioxide partial pressure, indicating that by using a unified time base, window mapping, and recursive fusion processing, the influence of asynchronous sampling and sensor fluctuations on the calculation results of carbonate system parameters can be reduced, thereby improving the stability and reliability of parameter reconstruction.

[0081] The above specific embodiments are used to explain and illustrate the present invention, but not to limit the present invention. Any modifications and changes made to the present invention within the spirit and scope of the claims shall fall within the protection scope of the present invention.

Claims

1. A multi-parameter carbonate system determination system, characterized in that, include: The liquid circuit unit is used to accurately deliver the sample mixture to be measured to the reaction cell (3) based on the conical flow guide and simultaneously recover the liquid; The gas path unit includes a drying-optimized condensation and purification module and a switchable carrier gas module. The carrier gas module and the liquid path unit are connected to the bottom of the reaction cell (3). The sample mixing system and the gas output from the carrier gas module are mixed to obtain the reaction product. The condensation and purification module is connected to the top of the reaction cell (3) to jointly complete the extraction and purification of the reaction product and recover the gas at the same time. A sensor detection unit is installed in the liquid circuit unit and the condensation and purification module to collect initial data of the carbonate system; The control unit (11) is electrically connected to the liquid circuit unit, the carrier gas module and the condensation and purification module. It is used to receive the initial data of the carbonate system, perform time window matching and recursive fusion processing and then obtain the parameters of the carbonate system to realize the determination of multi-parameter carbonate system.

2. The multi-parameter carbonate system determination system according to claim 1, characterized in that: The sensor detection unit includes an infrared detector (12), a pH sensor (15), and a temperature and salinity probe (16). The infrared detector (12) is installed in the condensation and purification module to collect CO2 concentration change data of the gas flowing through during the measurement cycle, and then inputs it into the control unit to obtain the dissolved inorganic carbon (DIC) result through integration. The pH sensor (15) and the temperature and salinity probe (16) are installed in the liquid circuit unit to collect pH, temperature and salinity data of the water body flowing through during the measurement period. The sampling frequency of the pH sensor (15) and the temperature and salinity probe (16) is higher than the data update frequency of the dissolved inorganic carbon (DIC) results. The infrared detector (12), pH sensor (15) and temperature and salinity probe (16) are all electrically connected to the control unit (11).

3. The multi-parameter carbonate system determination system according to claim 2, characterized in that: The middle part of the reaction tank (3) is a transparent cylinder. The top and bottom of the tank are provided with coaxial conical guide ports that pass through the middle part, which are connected to the condensation and purification module and the liquid circuit unit respectively. The diameter of the two conical guide ports closer to the middle part of the reaction tank (3) is larger than that of the side farther away from the middle part. A through hole is opened on the side of the bottom of the tank to connect the gas circuit unit.

4. The multi-parameter carbonate system determination system according to claim 3, characterized in that: The liquid circuit unit includes an injection pump (2), a waste liquid tank (13), a water pump (17), and a sample cell. The measurement is divided into laboratory mode and mobile mode. In laboratory mode, the water sample to be measured is directly stored in the sample cell. In mobile mode, the water sample to be measured is extracted from the mobile environment by the water pump (17). The pH sensor (15) and the temperature and salinity probe (16) are installed in the sample cell to measure the pH value, temperature and salinity data of the water sample. Each transmission port of the liquid inlet area (1) of the injection pump (2) is connected to the sample cell and the container storing phosphoric acid solution and distilled water through pipes. The sample mixing system consists of the water sample to be measured, phosphoric acid solution and distilled water. Another transmission port of the injection pump (2) is connected to the conical guide port at the bottom of the reaction tank (3). The output port of the injection pump (2) is connected to the waste liquid tank (13) to recover the liquid in the injection pump (2). The injection pump (2) and the water pump (17) are both electrically connected to the control unit (11).

5. The multi-parameter carbonate system determination system according to claim 3, characterized in that: The carrier gas module includes a carrier gas device (8), a second drying tube (9), and a flow meter (10). In laboratory mode, the carrier gas device (8) uses a nitrogen cylinder as the gas source. The nitrogen cylinder, the second drying tube (9), and the bottom holes of the reaction tank (3) are connected in sequence. The flow meter (10) is installed on the pipe between the second drying tube (9) and the bottom of the reaction tank (3). In mobile mode, the carrier gas device (8) uses an air pump as the gas source. The air pump is directly connected to the bottom of the reaction tank (3). The flow meter (10) is installed on the pipe between the air pump and the bottom of the reaction tank (3). The flow meter (10) controls the gas to flow at a preset speed by detecting the gas flowing through it. The sample mixing system and the nitrogen or air output by the carrier gas module are mixed to obtain the reaction product. The air pump and the flow meter (10) are both electrically connected to the control unit (11).

6. The multi-parameter carbonate system determination system according to claim 3, characterized in that: The condensation purification module includes a condenser (4), a first drying tube (5), a filter (6), a three-way valve (7), and a waste gas bottle (14) connected in sequence. The first outlet of the three-way valve (7) is directly connected to the waste gas bottle (14), and the second outlet of the three-way valve (7) is connected to the waste gas bottle (14) via an infrared detector (12). The condenser (4) is electrically connected to a power source, and the three-way valve (7) is electrically connected to a control unit (11).

7. The multi-parameter carbonate system determination system according to claim 6, characterized in that: The interior of the first drying tube (5) is a conical flow guide space that is thin at both ends and thick in the middle, in order to improve dehumidification efficiency and airflow stability.

8. The determination method of the multi-parameter carbonate system determination system according to any one of claims 1-7, characterized in that, include: Step 1) During the measurement period, the water sample to be measured is extracted multiple times. The control unit (11) receives the initial data of the carbonate system, including the CO2 concentration change data of the reaction products and the pH value, temperature and salinity data of the water sample to be measured. The pH value, temperature and salinity data of the water sample to be measured are all used as the initial measurement data. After the initial measurement data is recursively fused, the pre-processed measurement data is obtained. At the same time, the dissolved inorganic carbon (DIC) result is obtained based on the CO2 concentration change data. Step 2) Based on the dissolved inorganic carbon (DIC) results, determine several reaction time windows. For each pre-processed measurement data and each reaction time window, extract and match the data of the pre-processed measurement data within the reaction time window to the reaction time window. Then, perform weighted statistical processing to obtain the parameter characterization values ​​that characterize the water state of the water sample to be measured. Construct a multi-parameter synchronous data set by combining the dissolved inorganic carbon (DIC) results and the parameter characterization values ​​within each reaction time window. Step 3) Input the multi-parameter synchronous data set into the carbonate system calculation model, process it, and output the carbonate system parameters.

9. The determination method of the multi-parameter carbonate system determination system according to claim 8, characterized in that: In step 1), the recursive fusion processing specifically employs an estimation method based on state prediction and observation update to iteratively process the initial measurement data. Each iteration outputs a predicted value, and the residual between the predicted value and the initial measurement data is used as the dynamic residual of the estimation method based on state prediction and observation update. The process noise covariance and measurement noise covariance are dynamically adjusted based on the dynamic residual to adaptively correct the initial measurement data and obtain preprocessed measurement data.

10. The determination method of the multi-parameter carbonate system determination system according to claim 8, characterized in that: In step 2), the starting point of the reaction time window is the moment when the CO2 concentration in the dissolved inorganic carbon (DIC) results exceeds the baseline range, and the ending point is the moment when the CO2 concentration returns to the baseline range. During weighted statistical processing, for each pre-processed measurement data within each reaction time window, the weighted mean of the pre-processed measurement data is obtained. The data weights from the starting point to the ending point of the reaction time window in the pre-processed measurement data gradually decrease. Finally, the weighted mean of each pre-processed measurement data in each reaction time window is averaged and used as the parameter characterization value.

Images