A total phosphorus multi-parameter synchronous detection method based on water body
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- LUOYANG LAIBOTU ELECTRONIC TECH CO LTD
- Filing Date
- 2026-03-03
- Publication Date
- 2026-05-29
Smart Images
Figure CN121783968B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of water quality testing and analysis technology, and in particular to a method for simultaneous detection of multiple parameters of total phosphorus in water. Background Technology
[0002] In the detection of total phosphorus in water, conventional techniques typically employ a single chromogenic reagent optimized for a specific concentration range, measuring absorbance at a fixed time point after the reaction reaches equilibrium. For actual water samples with complex compositions, significant turbidity and color interference, or drastic fluctuations in total phosphorus concentration, single measurements are easily affected by both background interference and the nonlinear range of the chromogenic reaction, leading to decreased accuracy. To overcome interference, existing methods require cumbersome sample pretreatment or the setting of multiple physical blanks, resulting in complex procedures and difficulty in completely eliminating dynamic optical interference coexisting with the analyte. Furthermore, relying on a single chromogenic reaction and endpoint measurement mode fails to effectively acquire information about the reaction process, often leading to limited measurement range or misinterpretation of results when dealing with samples of unknown concentrations or in the presence of competing interfering substances.
[0003] The shortcomings of existing technologies lie in their serial and isolated measurement paradigms. The separation of digestion and optical measurement steps makes it impossible to effectively trace and subtract time-varying background spectral interference from the sample itself during measurement. Relying on a single chromogenic reagent and static measurement points makes it difficult to adapt to a wide concentration range and lacks the ability to identify and resist interference from complex matrices using reaction kinetic information. Therefore, a new method for total phosphorus detection in water is needed that can simultaneously acquire sample background characteristics, concurrently utilize multi-reaction path kinetic information, and directly analyze the final result through data fusion. This would enable rapid, stable, and high-precision determination of total phosphorus content in various water samples without the need for complex pretreatment. Summary of the Invention
[0004] The purpose of this invention is to overcome the shortcomings of existing technologies and propose a method for simultaneous detection of multiple parameters of total phosphorus in water.
[0005] To achieve the above objectives, the present invention adopts the following technical solution: a method for simultaneous detection of multiple parameters of total phosphorus in water, comprising:
[0006] Obtain the original water sample of the water body to be tested, and pretreat the original water sample to obtain a standard test solution; simultaneously introduce the standard test solution into a chemical digestion reaction cell and an optical feature extraction cell set in parallel.
[0007] In the chemical digestion reaction tank, a synchronous digestion program is initiated to convert different forms of phosphorus compounds in the standard test solution into a uniform form of phosphate.
[0008] In the optical feature extraction cell, in-situ spectral scanning is performed to capture the background spectral data of the standard test solution in the absence of added colorimetric reagent;
[0009] The intermediate liquid from the chemical digestion reaction tank was extracted after digestion and quantitatively distributed into multiple independent colorimetric reaction channels.
[0010] Different characteristic colorimetric agents are injected into each of the colorimetric reaction channels, and each characteristic colorimetric agent is configured for different concentration ranges of phosphate or interference conditions;
[0011] All the colorimetric reaction channels are controlled to carry out concurrent colorimetric reactions under the same environmental parameters. When the concurrent colorimetric reaction reaches a preset time node, the reaction solution in each colorimetric reaction channel is subjected to time-series spectral acquisition to obtain multiple reaction kinetic spectral sequences.
[0012] The background spectral data and all the reaction kinetic spectral sequences are input into the integrated calibration model for data fusion and cross-analysis. Based on the output of the integrated calibration model, the final detection result representing the total phosphorus content of the water body is analyzed and generated.
[0013] As a further aspect of the present invention, the pretreatment of the original water sample to obtain the standard test solution includes:
[0014] The original water sample is filtered to remove suspended particulate matter and large particulate impurities.
[0015] The filtered water sample is acidified and stabilized by adding a predetermined amount of acid stabilizer to adjust the pH of the water sample to a stable range and prevent the transformation or adsorption of phosphorus forms.
[0016] A quantitative dilution operation was performed on the acidified and stabilized water sample. The dilution factor was calculated according to the preset concentration prediction model, and the water sample was diluted with a phosphorus-free diluent to obtain the diluted water sample.
[0017] The diluted water sample was homogenized to ensure uniform component distribution, and then finally packaged as the standard test solution.
[0018] As a further aspect of the present invention, the step of initiating a simultaneous digestion process in the chemical digestion reaction tank to convert different forms of phosphorus compounds in the standard test solution into a uniform form of phosphate includes:
[0019] A mixed solution of oxidant and catalyst is injected into the chemical digestion reaction tank, wherein the oxidant is persulfate and the catalyst is an acid-base regulator;
[0020] The chemical digestion reaction tank after the injection of the mixed solution is started with a temperature rise program, so that the internal temperature rises to the target digestion temperature according to the preset temperature rise curve and is maintained.
[0021] During the target digestion temperature holding phase, periodic pressure pulsations are applied to the chemical digestion reaction tank to enhance the mass transfer efficiency of the digestion reaction.
[0022] The redox potential of the solution in the chemical digestion reaction tank is monitored by a built-in digestion process sensor. When the redox potential changes tend to stabilize, the synchronous digestion process is determined to be complete.
[0023] After digestion is complete, the solution in the chemical digestion reaction tank is cooled to room temperature to obtain the intermediate solution after digestion.
[0024] As a further aspect of the present invention, the step of performing in-situ spectral scanning in the optical feature extraction cell to capture the background spectral data of the standard test solution in the absence of added colorimetric reagent includes:
[0025] The temperature control unit in the optical feature extraction cell is controlled to ensure that the temperature of the standard test solution is consistent with the ambient temperature.
[0026] The broadband light source is activated to illuminate the standard test solution in the optical feature extraction cell;
[0027] The transmission spectrum through the standard test liquid and the scattering spectrum at a specific angle were simultaneously acquired using a fiber optic spectrometer.
[0028] The acquired transmission spectrum and scattering spectrum are spliced together and noise filtered to form a composite spectrum covering the ultraviolet, visible and part of the near-infrared bands.
[0029] The composite spectral data is stored as the background spectral data of the water body to be tested.
[0030] As a further aspect of the present invention, injecting different characteristic colorimetric agents into each of the colorimetric reaction channels includes:
[0031] Configure a set of characteristic colorimetric agents with different components, the characteristic colorimetric agents including molybdate-based main colorimetric agents, ascorbic acid-based reducing agents, and different types of masking agents or sensitizers;
[0032] Each of the colorimetric reaction channels is connected to an independent characteristic colorimetric reagent storage tank, and each characteristic colorimetric reagent storage tank stores a pre-configured characteristic colorimetric reagent.
[0033] According to the preset sample addition sequence, different volumes or different types of the characteristic colorimetric reagents are injected from their respective characteristic colorimetric reagent storage tanks into the corresponding colorimetric reaction channels through a precision liquid dispensing pump;
[0034] In each of the colorimetric reaction channels, the mixing ratio of the digested intermediate solution to the injected characteristic colorimetric agent is controlled by a pre-programmed addition scheme.
[0035] As a further aspect of the present invention, controlling all the colorimetric reaction channels to undergo concurrent colorimetric reactions under the same environmental parameters includes:
[0036] The reaction chamber containing all the colorimetric reaction channels is placed in a constant temperature environment to ensure that the reaction temperature of all channels is consistent and stable.
[0037] Apply micro-oscillation and stirring at the same frequency to the reaction solution in all the colorimetric reaction channels to ensure that the reactants are uniformly mixed and the reaction interface is constantly renewed;
[0038] Throughout the entire process of the concurrent colorimetric reaction, the ambient temperature, humidity, and light intensity within the reaction chamber are continuously monitored, and the environmental parameters are recorded as a reaction environment log.
[0039] As a further aspect of the present invention, the step of performing time-series spectral acquisition on the reaction solution in each of the colorimetric reaction channels to obtain multiple reaction kinetic spectral sequences includes:
[0040] Each colorimetric reaction channel is equipped with an independent spectral acquisition probe, and all spectral acquisition probes are connected to the same spectrometer through a multiplexer;
[0041] After the concurrent colorimetric reaction begins, the multiplexer cycles through the preset equal time intervals to transmit the spectral signals acquired by each spectral acquisition probe to the spectrometer in sequence.
[0042] During each switching cycle, the spectrometer records the absorbance or fluorescence intensity of the reaction solution in each colorimetric reaction channel within a specific wavelength range.
[0043] From the start of the reaction to the end of the preset time node, an optical signal data sequence that changes over time is generated for each of the colorimetric reaction channels, and each sequence is a reaction kinetic spectrum sequence.
[0044] As a further aspect of the present invention, the step of inputting the background spectral data and all the reaction kinetic spectral sequences into an integrated calibration model for data fusion and cross-analysis includes:
[0045] Read the background spectral data and extract characteristic spectral information reflecting water turbidity, color and background organic interference from the background spectral data;
[0046] Read the reaction kinetic spectral sequences of all the colorimetric reaction channels, and extract the final signal value of the colorimetric reaction reaching the plateau period, the characteristic value of the colorimetric reaction rate curve, and the duration of the reaction induction period from each sequence;
[0047] The characteristic spectral information reflecting the turbidity, color and background organic interference of the water body, together with the final signal value from different colorimetric reaction channels, the characteristic value of the colorimetric reaction rate curve and the reaction induction period, constitute a multidimensional feature vector.
[0048] The multidimensional feature vector is input into a pre-trained integrated calibration model, which is a neural network model.
[0049] Within the integrated calibration model, the multidimensional feature vector is processed sequentially through a feature weighting layer, a cross-channel information interaction layer, and a regression output layer.
[0050] As a further aspect of the present invention, the training process of the pre-trained integrated calibration model includes:
[0051] Using a large number of standard water samples with known total phosphorus concentrations as training samples, the entire process from preprocessing to time-series spectral acquisition was performed on each training sample to obtain the background spectral data set and reaction kinetic spectral sequence set corresponding to each training sample.
[0052] Based on the background spectral data set and the reaction kinetic spectral sequence set, a multidimensional feature vector for training is extracted and used to form a training data pair with the known total phosphorus concentration value.
[0053] The training data is used to initialize the integrated calibration model, and the weight parameters of the neural network model are adjusted by the backpropagation algorithm so that the output value of the model approximates the known total phosphorus concentration value.
[0054] During training, cross-validation is used to prevent overfitting until the model's prediction error on the validation set is stably lower than a preset threshold. At this point, the structure and parameters of the model are fixed and used as the pre-trained integrated calibration model.
[0055] As a further aspect of the present invention, the step of analyzing and generating a final detection result representing the total phosphorus content of the water body based on the output of the integrated calibration model includes:
[0056] Receive one or more numerical outputs generated by the regression output layer of the integrated calibration model;
[0057] The concentration conversion rules that are compatible with the integrated calibration model are invoked to map the one or more numerical outputs to a predicted value of total phosphorus concentration.
[0058] The estimated total phosphorus concentration was standardized in terms of unit and format.
[0059] The processed concentration value is marked as the final detection result, and a data record containing the sample identifier, detection timestamp, and the final detection result is generated.
[0060] Compared with the prior art, the advantages and positive effects of the present invention are as follows:
[0061] Homologous standard test solutions are injected in parallel into independent digestion and optical feature cells. Simultaneously with digestion, the full-spectral data of the water sample without any added chemical reagents is captured in situ and in real time. This background spectrum fully preserves the original optical properties of the sample within the detection wavelength range, including turbidity, color, and the absorption / scattering background of coexisting substances. In subsequent data processing, this spectrum can serve as a background benchmark completely homologous to the digested test solution, enabling the direct and accurate removal of optical background interference from the sample itself from the measurement signal, rather than relying on potentially erroneous theoretical calculations or the subtraction of blank samples from asynchronous measurements. This improves the accuracy and reliability of testing complex real-world water samples.
[0062] The digested intermediate solution was distributed to multiple independent channels configured with different characteristic chromogenic agents for concurrent reactions, and the time-series spectral sequences of the reaction solutions in each channel were acquired. Different chromogenic agents exhibit varying response characteristics, linear ranges, and anti-interference capabilities to phosphate, thus providing multi-dimensional reaction kinetic information. An integrated calibration model, by fusing and analyzing these kinetic spectral sequences from different reaction pathways, can comprehensively determine the optimal quantification range and identify and suppress specific substance interferences that may affect a single chromogenic agent. The concurrent measurement mode allows for the acquisition of information equivalent to multiple traditional experiments in a single analysis. Through cross-validation and collaborative analysis of the algorithm model, the effective detection range of the method is broadened, and its robustness to unknown or complex interferences is enhanced, ultimately achieving high-precision, adaptive concentration analysis. Attached Figure Description
[0063] Figure 1 This is a flowchart of the method for simultaneous detection of multiple parameters of total phosphorus in water as described in this invention;
[0064] Figure 2 Flowchart for the operation of a chemical digestion reaction tank;
[0065] Figure 3 The correlation curve between total phosphorus concentration and the prediction error of the integrated calibration model;
[0066] Figure 4Box plots showing the temperature stability of the reaction chamber at different time periods under constant temperature control at 30℃.
[0067] Figure 5 To compare the induction period duration of different colorimetric reaction channels across multiple samples. Detailed Implementation
[0068] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.
[0069] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.
[0070] See Figure 1 The process involves acquiring raw water samples from the target water body and pretreating them to obtain a standard test solution. This standard test solution is then simultaneously introduced into a parallel chemical digestion reaction chamber and an optical feature extraction chamber. In the chemical digestion reaction chamber, a synchronous digestion program is initiated to convert different forms of phosphorus compounds in the standard test solution into a uniform form of phosphate. In the optical feature extraction chamber, in-situ spectral scanning is performed to capture the background spectral data of the standard test solution without the addition of a colorimetric reagent. Subsequently, the digested intermediate solution is extracted from the chemical digestion reaction chamber and quantitatively distributed into multiple independent colorimetric reaction channels. Different characteristic colorimetric reagents are injected into each colorimetric reaction channel, each configured for different concentration ranges of phosphate or interference conditions. All colorimetric reaction channels are controlled to undergo concurrent colorimetric reactions under the same environmental parameters. When the concurrent colorimetric reactions reach a preset time point, time-series spectral acquisition is performed on the reaction solution in each colorimetric reaction channel to obtain multiple reaction kinetic spectral sequences. Finally, the background spectral data and all reaction kinetic spectral sequences are input into the integrated calibration model for data fusion and cross-analysis. Based on the output of the integrated calibration model, the final detection results representing the total phosphorus content of the water body are analyzed and generated.
[0071] In one embodiment of the present invention, the pretreatment stage of the method for simultaneous detection of total phosphorus in water involves processing the raw water sample to obtain a standard test solution. For example, a raw water sample collected from a lake may contain suspended particulate matter and large particulate impurities after a pretreatment process. The first step is to perform a filtration operation, using a filter membrane with a pore size of 0.45 micrometers to filter the raw water sample, removing the suspended particulate matter and large particulate impurities. The filtered water sample is clear. In some embodiments, for raw river water samples with high turbidity, the filtration operation can employ a multi-layer filter cartridge series method, first performing coarse filtration to remove large particulate impurities, and then performing fine filtration to remove fine particulate matter, thereby ensuring the stability of subsequent steps. The acidification stabilization operation is performed after filtration, adding a predetermined amount of acidic stabilizer, which is a dilute sulfuric acid solution, to the filtered water sample. The amount added is determined according to the water sample volume and initial pH, adjusting the pH of the water sample to a stable range, for example, adjusting the pH to below 2, to prevent phosphorus transformation or adsorption. In practice, the addition of acid stabilizer is controlled by an automatic titrator to ensure precise addition. After acidification and stabilization, the water sample is transferred to a sealed container to prevent air contact. It is understandable that the type and amount of acid stabilizer may be adjusted for raw water samples from different sources, but the core objective is to maintain the stability of phosphorus speciation.
[0072] The quantitative dilution operation calculates the dilution factor based on a pre-defined concentration prediction model. This model provides an initial concentration estimate based on historical data of the original water sample's source or rapid screening results; for example, it might use the average total phosphorus concentration of lake samples from historical monitoring data as the initial estimate. Optionally, the concentration prediction model can calculate the dilution factor based on a formula, expressed as:
[0073]
[0074] in: Represents the dilution factor. The estimated total phosphorus concentration represents the original water sample. The target total phosphorus concentration represents the standard test solution and is set as the midpoint of the linear range of the detection method. In practice, a phosphorus-free diluent, deionized water, is used for dilution. The dilution process is carried out in a clean environment to avoid contamination, resulting in a diluted water sample. Data comparison reflects the differences in dilution factors for different original water samples. For example, for industrial wastewater samples with a high estimated total phosphorus concentration, the dilution factor may be larger, while for drinking water samples with a low estimated total phosphorus concentration, the dilution factor may be smaller. However, all operations follow the same calculation logic.
[0075] Homogenization is performed on the diluted water sample. The diluted water sample is placed on a vortex mixer and vortexed to ensure uniform component distribution. The mixing time is set to at least 30 seconds, and then the sample is packaged in a standard sample vial as a standard test solution. In some embodiments, homogenization can also be performed using ultrasonic stirring to enhance homogenization. The packaged standard test solution is labeled with the sample identifier and stored in a light-protected environment for subsequent testing. It can be understood that the entire pretreatment process, from filtration to homogenization, forms a coherent sequence, with the output of each step serving as the input for the next step, ultimately yielding a standardized liquid that meets the testing requirements.
[0076] See Figure 2 In one embodiment of the present invention, the simultaneous digestion process in the chemical digestion reactor aims to convert different forms of phosphorus compounds in the standard test solution into a uniform form of phosphate. For example, this process treats standard test solutions from the effluent of a municipal wastewater treatment plant, which may contain multiple forms such as organophosphates and polyphosphates. The first step involves injecting a mixed solution of oxidant and catalyst into the chemical digestion reactor. The oxidant is a potassium persulfate solution, and the catalyst is a sodium hydroxide solution. The injection volume is determined based on the volume of the standard test solution and the target pH. In some embodiments, for industrial wastewater standard test solutions containing recalcitrant organophosphate compounds, the concentration and ratio of the oxidant and catalyst mixed solution can be adjusted, but the oxidant remains persulfate, and the catalyst remains an acid-base regulator to maintain the alkaline environment required for digestion.
[0077] The programmed temperature rise begins after the mixed solution is injected. The internal temperature of the chemical digestion reaction tank increases according to a preset temperature rise curve, defined as a function of temperature over time, for example, rising from room temperature to the target digestion temperature of 120 degrees Celsius at a rate of 5 degrees Celsius per minute and maintaining that temperature. In practice, the temperature rise process is executed by an embedded temperature control system within the reaction tank. The system adjusts the heating power through feedback to ensure that the temperature trajectory matches the preset curve. It is understandable that different water matrices may correspond to different temperature rise curves. For standard test solutions with complex compositions, the temperature rise curve may include multiple temperature plateaus to promote the gradual conversion of different phosphorus forms, but the core objective is to reach and maintain the target digestion temperature.
[0078] During the target digestion temperature holding phase, periodic pressure pulsations are applied to the chemical digestion reaction tank. These pulsations are achieved via a miniature pneumatic pump connected to the top of the tank to enhance the mass transfer efficiency of the digestion reaction. For example, the pressure pulsation mode is set to complete a cycle from atmospheric pressure to 150 kPa and back once per second. This periodic compression and release of the perturbed solution promotes collisions of reactant molecules. Data comparisons demonstrate the effects under different pressure pulsation parameters. For high-viscosity standard test solutions or those containing colloidal substances, the amplitude and frequency of the pressure pulsations can be increased, while for clear, low-turbidity standard test solutions, baseline parameters can be used. However, all operations are performed during the target digestion temperature holding phase. Optionally, the application of pressure pulsations can be associated with an empirical formula describing enhanced mass transfer, expressed as:
[0079]
[0080] in: The mass transfer efficiency factor per unit volume. The angular frequency representing pressure pulsations. This represents the amplitude of the pressure pulsation. This represents the volume of the solution within the chemical digestion reaction tank.
[0081] The redox potential (RPP) of the solution in the chemical digestion reaction tank is monitored by a built-in digestion process sensor, which is a combination of platinum electrode and reference electrode, to measure the RRP value of the solution in real time. In specific implementation, the RRP value gradually shifts from a high negative value or fluctuating state at the beginning of digestion to a stable plateau as digestion progresses. When the fluctuation range of the RRP change is less than 5 mV over 60 consecutive seconds, the synchronous digestion process is considered complete. In some embodiments, the digestion process sensor can also monitor the conductivity of the solution as an auxiliary criterion, but the change in RRP is the primary criterion. After digestion, the cooling module of the chemical digestion reaction tank is activated to cool the solution in the tank to room temperature at a controllable rate, for example, to 25 degrees Celsius at a rate not exceeding 10 degrees Celsius per minute, to obtain the intermediate solution after digestion. This intermediate solution should be clear and all phosphorus forms have been converted to orthophosphate. It is understood that the cooling process must avoid drastic temperature changes that could alter the solution composition. The cooled intermediate solution is immediately transferred or enters the next stage to prevent precipitation.
[0082] In one embodiment of the invention, the optical feature extraction cell is operated to capture the background spectral data of a standard test solution in its undeveloped state, such as processing a standard test solution containing natural humic acid and fine clay particles from a river. A temperature control unit within the optical feature extraction cell is used to maintain the temperature of the standard test solution at the ambient temperature, for example, precisely maintaining the temperature of the standard test solution within the optical feature extraction cell at 25 degrees Celsius. A broadband light source is activated to illuminate the standard test solution within the optical feature extraction cell; the broadband light source has an emission wavelength range covering 200 nm to 900 nm. A fiber optic spectrometer is used to simultaneously acquire the transmission spectrum transmitted through the standard test solution, as well as the scattering spectrum at a specific angle set at 90 degrees to the incident light path.
[0083] The acquired transmission and scattering spectra are stitched together and noise filtered. The noise filtering employs the Savitzky-Golay smoothing algorithm to create a composite spectrum covering the ultraviolet, visible, and part of the near-infrared bands. In practice, spectral stitching involves weighting and superimposing spectral signals from different bands. For example, for high-turbidity standard test solutions, the weighting coefficient of the scattering spectral signal in the composite spectrum is increased. The generation process of the composite spectrum can be understood as following a linear combination formula, expressed as:
[0084]
[0085] in: This represents the final generated composite spectral data vector. This represents the vector of acquired transmission spectrum data. This represents a vector of scattering spectrum data collected at a specific angle. and These are weighting coefficients dynamically adjusted based on the estimated turbidity of the water sample. Data comparison is reflected in the processing of different water quality samples. For lake water standard test solutions rich in color but low in turbidity, the weighting coefficients... The value is greater than the weighting coefficient. For high-turbidity estuarine standard test solutions, the weighting coefficient... The value is adjusted accordingly. The composite spectral data is stored as the background spectral data of the water body to be tested. The background spectral data is stored in the form of a digital matrix, where the columns of the matrix correspond to the wavelength and the rows correspond to the light intensity values.
[0086] A set of characteristic chromogenic reagents with different components is configured. These reagents include a molybdate-based main chromogenic reagent, an ascorbic acid-based reducing agent, and various types of masking agents or sensitizers. For example, a set of characteristic chromogenic reagents may include three formulations: the first formulation adds potassium antimony tartrate as a sensitizer to adapt to the low concentration range; the second formulation adds EDTA as a masking agent to counteract iron ion interference; and the third formulation adjusts the relative ratio of ammonium molybdate to ascorbic acid to expand the high concentration detection range. Each chromogenic reaction channel is connected to an independent characteristic chromogenic reagent storage tank. Each tank stores one pre-configured characteristic chromogenic reagent and is equipped with light-proof and inert gas protection functions. In some embodiments, the types of characteristic chromogenic reagents can be preset according to common interference types in water bodies. For example, for seawater samples with high salinity, the characteristic chromogenic reagent may include a masking component for chloride ions, but all characteristic chromogenic reagents are based on a molybdate and ascorbic acid chromogenic system.
[0087] According to a preset sample addition sequence, a precision liquid dispensing pump injects different volumes or types of characteristic colorimetric reagents from their respective characteristic colorimetric reagent tanks into the corresponding colorimetric reaction channels. For example, for a system configured with four independent colorimetric reaction channels, the sample addition sequence instructs the precision liquid dispensing pump to inject 0.5 ml of the first type of characteristic colorimetric reagent into the first channel at time zero, and 0.8 ml of the second type of characteristic colorimetric reagent into the second channel one second after time zero, and so on. Optionally, the sample addition sequence can be dynamically adjusted according to the estimated total phosphorus concentration of the digested intermediate solution, but all injection actions are automatically completed by program control. In each colorimetric reaction channel, the mixing ratio of the digested intermediate solution to the injected characteristic colorimetric reagent is controlled by a pre-programmed sample addition scheme. For example, in the first colorimetric reaction channel, the mixing ratio is set to a ratio of intermediate solution volume to characteristic colorimetric reagent volume of 10:1, while in the second colorimetric reaction channel, this ratio is adjusted to 10:1.5. It is understood that different mixing ratios are intended to bring the colorimetric reaction of each channel into its optimal detection kinetic range. The mixing ratio is a fixed parameter in the sample addition scheme or a variable adjusted based on feedback. In some embodiments, the injection accuracy of the precision liquid dispensing pump needs to reach micro-levels to ensure the reproducibility of the mixing ratio, and all characteristic colorimetric reagent injection operations are performed under isothermal conditions to avoid temperature-introduced errors.
[0088] See Figure 3In the performance validation of the integrated calibration model, the correlation analysis between total phosphorus concentration and prediction error revealed the model's performance characteristics across different concentration ranges. Specifically, the prediction error was defined as the absolute deviation between the integrated calibration model's output value and the actual total phosphorus concentration value, expressed in mg / L. In the figure, the solid line represents the average prediction error at different total phosphorus concentrations, the light blue area represents the 95% confidence interval of the error, and the dashed line represents the preset maximum permissible error threshold (0.008 mg / L). The curve trend shows that when the total phosphorus concentration is in the range of 0.0–0.5 mg / L, the prediction error increases linearly with increasing concentration; it reaches a peak at 0.5 mg / L (approximately 0.008 mg / L), which just touches the maximum permissible error threshold; subsequently, in the range of 0.5–1.2 mg / L, the prediction error decreases with increasing concentration. This variation reflects the fitting pressure of the integrated calibration model in the medium concentration range, and its stability advantage in the low and high concentration ranges. In terms of parameter configuration, the maximum permissible error threshold is set based on the accuracy requirements of total phosphorus determination in water quality testing standards, while the calculation of the confidence interval depends on the error sample distribution of 100 repeated experiments.
[0089] In one embodiment of the invention, controlling all colorimetric reaction channels to undergo concurrent colorimetric reactions under the same environmental parameters involves placing the reaction chamber containing all colorimetric reaction channels in a constant-temperature environment, such as by wrapping the reaction chamber with a constant-temperature circulating water bath, to ensure that the reaction temperature of all colorimetric reaction channels is consistent and stable at 30 degrees Celsius. Micro-oscillation stirring at the same frequency is applied to the reaction liquid in all colorimetric reaction channels. This micro-oscillation stirring is achieved by a piezoelectric ceramic actuator installed at the bottom of the reaction chamber, which generates vertical micro-vibrations at a frequency of 50 Hz to ensure uniform mixing of the reactants and continuous renewal of the reaction interface. Throughout the concurrent colorimetric reaction, the ambient temperature, humidity, and light intensity within the reaction chamber are continuously monitored. The ambient temperature is measured using a platinum resistance temperature sensor, the humidity using a capacitive humidity sensor, and the light intensity using a silicon photodetector. The environmental parameters are recorded as a reaction environment log, which is stored in the form of a time-series data file.
[0090] In some embodiments, a Peltier thermoelectric temperature control module can be used instead of a circulating water bath to establish a constant temperature environment. The Peltier thermoelectric temperature control module is directly attached to the outer wall of the reaction chamber for bidirectional temperature control, but the core purpose remains to ensure that the reaction temperature of all colorimetric reaction channels is consistent and stable. The frequency of micro-oscillation stirring can be adjusted according to the viscosity of the reaction solution. For reaction solutions containing thickeners or colloids, the frequency of micro-oscillation stirring can be increased to 100 Hz to enhance the mixing effect, but the frequency of micro-oscillation stirring applied to all colorimetric reaction channels is always kept the same. In specific implementations, the recording of the reaction environment log is synchronized with the time sequence of concurrent colorimetric reactions. The log entries include timestamps, temperature readings, humidity readings, and light intensity readings for subsequent data correlation analysis.
[0091] Time-series spectral acquisition is performed on the reaction solution in each colorimetric reaction channel to obtain multiple reaction kinetic spectral sequences. In specific implementation, each colorimetric reaction channel is equipped with an independent spectral acquisition probe, which consists of an optical fiber bundle and a focusing lens. All spectral acquisition probes are connected to the same spectrometer through a multiplexer. After the concurrent colorimetric reaction begins, the multiplexer cycles through the channels at preset equal time intervals, allowing each spectral acquisition probe to sequentially transmit its acquired spectral signal to the spectrometer. For example, the preset equal time interval is set to 2 seconds. The multiplexer completes one polling scan of all colorimetric reaction channels within each interval. During each switching cycle, the spectrometer records the absorbance or fluorescence intensity values of the reaction solution in each colorimetric reaction channel within a specific wavelength range. The specific wavelength range is set to 700 nm to 900 nm to monitor the absorption characteristics of the phosphomolybdic blue colorimetric product.
[0092] From the start of the reaction to the end of the preset time point, a time-varying optical signal data sequence is generated for each colorimetric reaction channel, and each sequence is a reaction kinetic spectrum sequence. Data comparison is reflected in the acquisition parameters of different colorimetric reaction channels. Refer to Table 1, which shows an example of a time-series spectral acquisition configuration for a system with four colorimetric reaction channels.
[0093] Table 1: Time-series spectral acquisition parameters for colorimetric reaction channels
[0094]
[0095] It is understandable that the switching logic of the multiplexer follows a fixed polling algorithm, which ensures that each spectral acquisition probe is accessed within an equal time window. Optionally, the equal time interval can be dynamically calculated based on a formula, expressed as:
[0096]
[0097] in: This represents the time interval at which the multiplexer resides on each spectral acquisition probe. This represents the total data collection time from the start of the reaction to the preset time point. This represents the number of channels in the colorimetric reaction. This represents the number of full-channel scan cycles required to be completed within the total duration. In some embodiments, the spectrometer's recording mode can be a continuous integration mode, where the spectrometer integrates the signal from the currently active spectral acquisition probe in each switching cycle to improve the signal-to-noise ratio, but the core steps remain cyclic switching and sequential recording.
[0098] The time-series spectral acquisition process is synchronized with the environmental control of the concurrent colorimetric reaction. The timestamps in the reaction environment log are aligned with the timestamps of the spectral acquisition to ensure that each reaction kinetic spectral sequence point can be associated with the corresponding environmental parameters. It can be understood that the choice of spectral acquisition probe type depends on the optical design of the colorimetric reaction channel; transmission probes are used to measure absorbance, and reflection probes are used to measure the diffuse reflectance signal of the turbid reaction solution. However, all spectral acquisition probes are connected to the spectrometer through the same multiplexer for centralized acquisition. In practice, the generated reaction kinetic spectral sequences are stored in matrix form, with rows corresponding to acquisition time points and columns corresponding to wavelength points. Each colorimetric reaction channel generates an independent matrix file.
[0099] See Figure 4 In the colorimetric reaction stage of simultaneous multi-parameter detection of total phosphorus in water, the temperature stability distribution characteristics of the reaction chamber under a preset constant temperature control of 30℃ were demonstrated across different time intervals. Specifically, the upper and lower quartiles (Q1, Q3) and the median of the chamber visually represent the central tendency and dispersion of temperature at different time intervals: the median temperature for all time intervals fluctuates around the target temperature of 30℃, indicating that the overall temperature control accuracy of the system meets the standard. The height of the chamber (interquartile range, IQR) reflects the temperature fluctuation amplitude within each time interval. The chamber height is largest during the 0-50 second interval, indicating slightly weaker temperature stability at the initial stage of system startup. The chamber height gradually narrows in subsequent intervals, reflecting the dynamic response and stability capability of the temperature control system. Outliers (circles outside the upper and lower quartiles) reveal extreme temperature fluctuations within each time interval, such as the low-temperature anomaly during the 50-100 second interval and the high-temperature anomaly during the 250-300 second interval. These fluctuations may be related to the thermal response delay of the Peltier thermoelectric temperature control module or the flow rate fluctuations of the constant-temperature circulating water bath. From a temporal perspective, as the reaction progresses, the length of the upper and lower sections of the chamber gradually shortens, indicating that the temperature fluctuation range of the reaction chamber gradually converges, verifying the auxiliary effect of micro-amplitude oscillation stirring (50Hz) on temperature control uniformity. The figure provides crucial evidence of temperature stability for subsequent correlation analysis of the reaction environment log and reaction kinetic spectral sequence, which can be used to quantify the impact of temperature fluctuations on the rate and absorbance of the phosphomolybdic blue colorimetric reaction.
[0100] In one embodiment of the present invention, inputting background spectral data and all reaction kinetic spectral sequences into an integrated calibration model for data fusion and cross-analysis involves reading background spectral data and extracting characteristic spectral information reflecting water turbidity, color, and background organic interference from the background spectral data. For example, when processing samples from secondary effluent of urban wastewater treatment plants, absorption peaks in the 400-450 nm band of background spectral data may indicate interference from humic acid-like background organic matter, and the scattering background intensity in the 700-900 nm band may reflect water turbidity. The reaction kinetic spectral sequences of all colorimetric reaction channels were read, and the final signal value at which the colorimetric reaction reaches the plateau, the characteristic value of the colorimetric reaction rate curve, and the duration of the reaction induction period were extracted from each sequence. The final signal value at which the colorimetric reaction reaches the plateau was the average of the absorbance readings from 300 seconds to 360 seconds after the start of the reaction. The characteristic value of the colorimetric reaction rate curve was obtained by calculating the fitted slope of the absorbance-time curve in the maximum slope segment. The duration of the reaction induction period was defined as the time from the start of the colorimetric reaction to the point when the absorbance first exceeds three standard deviations of the background noise threshold.
[0101] The characteristic spectral information reflecting water turbidity, color, and background organic interference, along with the final signal values from different colorimetric reaction channels, the characteristic values of the colorimetric reaction rate curve, and the reaction induction period duration, are combined to form a multidimensional feature vector. In practice, the construction of the multidimensional feature vector follows a fixed dimensional order. For example, the first to fifth dimensions contain the average intensity values extracted from the five characteristic spectral bands of the background spectrum; the sixth to tenth dimensions contain parameters such as the final signal value of the first colorimetric reaction channel, the characteristic value of the colorimetric reaction rate curve, and the reaction induction period duration; subsequent dimensions contain the corresponding parameters of other colorimetric reaction channels. The multidimensional feature vector is input into a pre-trained ensemble calibration model, which is a neural network model using a feedforward fully connected structure. Inside the ensemble calibration model, the multidimensional feature vector is processed sequentially through a feature weighting layer, a cross-channel information interaction layer, and a regression output layer. The feature weighting layer assigns a learnable weight coefficient to each input feature; the cross-channel information interaction layer achieves nonlinear combinations of feature parameters extracted from different colorimetric reaction channels through neuronal connections; and the regression output layer ultimately produces one or more numerical outputs. Data comparison is reflected in the differences in multidimensional feature vectors when processing different water samples. For example, for river water samples with high turbidity and high chromaticity, the spectral information values reflecting turbidity and chromaticity in their multidimensional feature vectors are relatively high, while the reaction induction period duration parameter extracted from the reaction channel containing masking agent characteristic chromogenic agent may be relatively short.
[0102] The training process of the pre-trained integrated calibration model uses a large number of standard water samples with known total phosphorus concentrations as training samples. These standard water samples are prepared by dilution or formulation of nationally certified standard materials, covering different concentration ranges from low to high. A complete workflow from preprocessing to time-series spectral acquisition is performed on each training sample to obtain the corresponding background spectral data set and reaction kinetic spectral sequence set. Based on the background spectral data set and reaction kinetic spectral sequence set, multi-dimensional feature vectors for training are extracted according to the aforementioned method, and these vectors are paired with known total phosphorus concentration values to form training data pairs. The integrated calibration model is initialized using these training data pairs. The weight parameters of the neural network model are adjusted using a backpropagation algorithm to make the model's output value approximate the known total phosphorus concentration value. The loss function used in the backpropagation algorithm is the mean squared error function. During training, cross-validation is used to prevent overfitting. For example, all training data is randomly divided into five mutually exclusive subsets. Four subsets are used for training in turn, and the remaining subset is used for validation until the model's prediction error on the validation set is stably lower than a preset threshold. At this point, the model's structure and parameters are fixed, and this is used as the pre-trained integrated calibration model. It is understandable that the convergence of the training process can be judged by monitoring the changes in the value of the validation set loss function. Training should be stopped when the value of the validation set loss function no longer decreases significantly over multiple consecutive training cycles.
[0103] In some embodiments, the architectural details of the neural network model can be adjusted; for example, the feature weighting layer can be replaced with an attention mechanism layer to dynamically adjust feature importance. However, the basic training process of the integrated calibration model—namely, acquiring data from standard water samples, extracting features, forming training pairs, adjusting parameters through backpropagation, and using cross-validation to prevent overfitting—remains unchanged. Optionally, a regularization term can be introduced into the calculation of the loss function to further improve the model's generalization ability. The calculation of the regularization term is based on the norm of the model weight parameters. For example, the form of the loss function is:
[0104]
[0105] in: This represents the total loss function value. This represents the number of samples in a training batch. Representing the Known total phosphorus concentration values for each training sample. Represents the integrated calibration model for the first The predicted output value of each training sample. Represents the regularization coefficient. This represents a vector composed of all weight parameters in the integrated calibration model. This represents the squared L2 norm of the weight vector.
[0106] According to the output of the integrated calibration model, parse and generate the final detection result representing the total phosphorus content in the water body. Receive one or more numerical outputs generated by the regression output layer of the integrated calibration model. Invoke the concentration conversion rules配套 with the integrated calibration model to map the one or more numerical outputs to the estimated value of the total phosphorus concentration. The concentration conversion rules are usually a linear or polynomial mapping function, and its coefficients are determined synchronously after the integrated calibration model is trained. Perform unit unification and format standardization on the estimated value of the total phosphorus concentration. For example, uniformly convert the estimated value to a numerical value in milligrams per liter and format it as a numeric string with three decimal places reserved. In specific implementations, the unit unification process may involve unit conversion, which needs to be performed when the output of the integrated calibration model is based on other concentration units. The format standardization process follows the data specification of the test report. Mark the processed concentration value as the final detection result and generate a data record containing the sample identifier, detection timestamp, and the final detection result. The data record is stored in the form of a structured text file or database entry. It can be understood that the sample identifier is consistent with the identifier marked in the preprocessing stage, the detection timestamp is accurate to the second level, and the final detection result is output as the conclusion of this measurement and analysis. In some embodiments, the regression output layer of the integrated calibration model may output multiple numerical values. For example, it may output the estimated value of the total phosphorus concentration and the corresponding confidence index simultaneously. The concentration conversion rules need to be able to process multiple outputs and generate a structured final detection result record.
[0107] See Figure 5 , which presents the differences in the induction period duration of the three chromogenic reaction channels in different samples. Specifically, the induction period duration of Channel 1 (molybdate main chromogenic agent) is generally in a relatively high range, reaching a peak (about 59.5 seconds) at Sample 008 and dropping to a relatively low point (about 28.5 seconds) at Sample 002, reflecting the stability fluctuations of this channel under different water sample matrices. The induction period duration of Channel 2 (ascorbic acid reducing agent) fluctuates most violently, showing obvious peaks (up to about 58.5 seconds) at Samples 001, 003, and 006, while dropping to a low level of 10 - 15 seconds at Samples 002 and 007. This is closely related to the fact that the reduction efficiency of ascorbic acid is affected by the concentration of oxidizing impurities in the water sample. The induction period duration of Channel 3 (masking agent sensitizer) is generally in the medium range of 20 - 40 seconds, with the lowest value (about 13 seconds) at Sample 005 and the highest value (about 40.8 seconds) at Sample 003, reflecting the dynamic balance of the inhibitory effect of the masking agent on interfering substances and the synergistic effect of the sensitizer in different samples.
[0108] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.
Claims
1. A method for simultaneous detection of multiple parameters of total phosphorus in water, characterized in that, The method includes: Obtain the original water sample of the water body to be tested, and pretreat the original water sample to obtain a standard test solution; simultaneously introduce the standard test solution into a chemical digestion reaction cell and an optical feature extraction cell set in parallel. In the chemical digestion reaction tank, a synchronous digestion program is initiated to convert different forms of phosphorus compounds in the standard test solution into a uniform form of phosphate. In the optical feature extraction cell, in-situ spectral scanning is performed to capture the background spectral data of the standard test solution in the absence of added colorimetric reagent; The intermediate liquid from the chemical digestion reaction tank was extracted after digestion and quantitatively distributed into multiple independent colorimetric reaction channels. Different characteristic colorimetric agents are injected into each of the colorimetric reaction channels, and each characteristic colorimetric agent is configured for different concentration ranges of phosphate or interference conditions; All the colorimetric reaction channels are controlled to carry out concurrent colorimetric reactions under the same environmental parameters. When the concurrent colorimetric reaction reaches a preset time node, the reaction solution in each colorimetric reaction channel is subjected to time-series spectral acquisition to obtain multiple reaction kinetic spectral sequences. The background spectral data and all the reaction kinetic spectral sequences are input into the integrated calibration model for data fusion and cross-analysis. Based on the output of the integrated calibration model, the final detection result representing the total phosphorus content of the water body is analyzed and generated. The step of inputting the background spectral data and all the reaction kinetic spectral sequences into the integrated calibration model for data fusion and cross-analysis includes: Read the background spectral data and extract characteristic spectral information reflecting water turbidity, color and background organic interference from the background spectral data; Read the reaction kinetic spectral sequences of all the colorimetric reaction channels, and extract the final signal value of the colorimetric reaction reaching the plateau period, the characteristic value of the colorimetric reaction rate curve, and the duration of the reaction induction period from each sequence; The characteristic spectral information reflecting the turbidity, color and background organic interference of the water body, together with the final signal value from different colorimetric reaction channels, the characteristic value of the colorimetric reaction rate curve and the reaction induction period, constitute a multidimensional feature vector. The multidimensional feature vector is input into a pre-trained integrated calibration model, which is a neural network model. Within the integrated calibration model, the multidimensional feature vector is processed sequentially through a feature weighting layer, a cross-channel information interaction layer, and a regression output layer; The training process of the pre-trained integrated calibration model includes: Using a large number of standard water samples with known total phosphorus concentrations as training samples, the entire process from preprocessing to time-series spectral acquisition was performed on each training sample to obtain the background spectral data set and reaction kinetic spectral sequence set corresponding to each training sample. Based on the background spectral data set and the reaction kinetic spectral sequence set, a multidimensional feature vector for training is extracted and used to form a training data pair with the known total phosphorus concentration value. The training data is used to initialize the integrated calibration model, and the weight parameters of the neural network model are adjusted by the backpropagation algorithm so that the output value of the model approximates the known total phosphorus concentration value. During training, cross-validation is used to prevent overfitting until the model's prediction error on the validation set is stably lower than a preset threshold. At this point, the structure and parameters of the model are fixed and used as the pre-trained integrated calibration model.
2. The method for simultaneous detection of multiple parameters of total phosphorus in water as described in claim 1, characterized in that, The pretreatment of the original water sample to obtain the standard test solution includes: The original water sample is filtered to remove suspended particulate matter and large particulate impurities. The filtered water sample is acidified and stabilized by adding a predetermined amount of acid stabilizer to adjust the pH of the water sample to a stable range and prevent the transformation or adsorption of phosphorus forms. A quantitative dilution operation was performed on the acidified and stabilized water sample. The dilution factor was calculated according to the preset concentration prediction model, and the water sample was diluted with a phosphorus-free diluent to obtain the diluted water sample. The diluted water sample was homogenized to ensure uniform component distribution, and then finally packaged as the standard test solution.
3. The method for simultaneous detection of multiple parameters of total phosphorus in water according to claim 2, characterized in that, In the chemical digestion reaction tank, a simultaneous digestion process is initiated to convert different forms of phosphorus compounds in the standard test solution into a uniform form of phosphate, including: A mixed solution of oxidant and catalyst is injected into the chemical digestion reaction tank, wherein the oxidant is persulfate and the catalyst is an acid-base regulator; The chemical digestion reaction tank after the injection of the mixed solution is started with a temperature rise program, so that the internal temperature rises to the target digestion temperature according to the preset temperature rise curve and is maintained. During the target digestion temperature holding phase, periodic pressure pulsations are applied to the chemical digestion reaction tank to enhance the mass transfer efficiency of the digestion reaction. The redox potential of the solution in the chemical digestion reaction tank is monitored by a built-in digestion process sensor. When the redox potential changes tend to stabilize, the synchronous digestion process is determined to be complete. After digestion is complete, the solution in the chemical digestion reaction tank is cooled to room temperature to obtain the intermediate solution after digestion.
4. The method for simultaneous detection of multiple parameters of total phosphorus in water according to claim 3, characterized in that, The process involves performing an in-situ spectral scan in the optical feature extraction cell to capture the background spectral data of the standard test solution in the absence of a colorimetric reagent, including: The temperature control unit in the optical feature extraction cell is controlled to ensure that the temperature of the standard test solution is consistent with the ambient temperature. The broadband light source is activated to illuminate the standard test solution in the optical feature extraction cell; The transmission spectrum through the standard test liquid and the scattering spectrum at a specific angle were simultaneously acquired using a fiber optic spectrometer. The acquired transmission spectrum and scattering spectrum are spliced together and noise filtered to form a composite spectrum covering the ultraviolet, visible and part of the near-infrared bands. The composite spectral data is stored as the background spectral data of the water body to be tested.
5. The method for simultaneous detection of multiple parameters of total phosphorus in water according to claim 4, characterized in that, The injection of different characteristic colorimetric agents into each of the colorimetric reaction channels includes: Configure a set of characteristic colorimetric agents with different components, the characteristic colorimetric agents including molybdate-based main colorimetric agents, ascorbic acid-based reducing agents, and different types of masking agents or sensitizers; Each of the colorimetric reaction channels is connected to an independent characteristic colorimetric reagent storage tank, and each characteristic colorimetric reagent storage tank stores a pre-configured characteristic colorimetric reagent. According to the preset sample addition sequence, different volumes or different types of the characteristic colorimetric reagents are injected from their respective characteristic colorimetric reagent storage tanks into the corresponding colorimetric reaction channels through a precision liquid dispensing pump; In each of the colorimetric reaction channels, the mixing ratio of the digested intermediate solution to the injected characteristic colorimetric agent is controlled by a pre-programmed addition scheme.
6. The method for simultaneous detection of multiple parameters of total phosphorus in water according to claim 5, characterized in that, The control of all the colorimetric reaction channels to perform concurrent colorimetric reactions under the same environmental parameters includes: The reaction chamber containing all the colorimetric reaction channels is placed in a constant temperature environment to ensure that the reaction temperature of all channels is consistent and stable. Apply micro-oscillation and stirring at the same frequency to the reaction solution in all the colorimetric reaction channels to ensure that the reactants are uniformly mixed and the reaction interface is constantly renewed; Throughout the entire process of the concurrent colorimetric reaction, the ambient temperature, humidity, and light intensity within the reaction chamber are continuously monitored, and the environmental parameters are recorded as a reaction environment log.
7. The method for simultaneous detection of multiple parameters of total phosphorus in water according to claim 6, characterized in that, The step involves performing time-series spectral acquisition on the reaction solution in each of the colorimetric reaction channels to obtain multiple reaction kinetic spectral sequences, including: Each colorimetric reaction channel is equipped with an independent spectral acquisition probe, and all spectral acquisition probes are connected to the same spectrometer through a multiplexer; After the concurrent colorimetric reaction begins, the multiplexer cycles through the preset equal time intervals to transmit the spectral signals acquired by each spectral acquisition probe to the spectrometer in sequence. During each switching cycle, the spectrometer records the absorbance or fluorescence intensity of the reaction solution in each colorimetric reaction channel within a specific wavelength range. From the start of the reaction to the end of the preset time node, an optical signal data sequence that changes over time is generated for each of the colorimetric reaction channels, and each sequence is a reaction kinetic spectrum sequence.
8. The method for simultaneous detection of multiple parameters of total phosphorus in water according to claim 7, characterized in that, The step of parsing and generating the final detection result representing the total phosphorus content of the water body based on the output of the integrated calibration model includes: Receive one or more numerical outputs generated by the regression output layer of the integrated calibration model; The concentration conversion rules that are compatible with the integrated calibration model are invoked to map the one or more numerical outputs to a predicted value of total phosphorus concentration. The estimated total phosphorus concentration was standardized in terms of unit and format. The processed concentration value is marked as the final detection result, and a data record containing the sample identifier, detection timestamp, and the final detection result is generated.