Detection Method of High Potassium Organic Water-Soluble Fertilizer Components Based on Spectroscopic Analysis

Abstract

This invention relates to the field of agricultural fertilizer detection and spectral analysis technology, specifically a method for detecting high-potassium organic water-soluble fertilizer components based on spectral analysis. The method is implemented on a detection device with a sample inlet flow path, a dilution branch, and long and short optical path flow cells. It includes: acquiring conductivity, temperature, pressure difference, bubble passage signal, reference light intensity, blank background spectrum, and sample spectrum; determining the transmittance load level, flow stability, and residual contamination level accordingly; selecting the corresponding detection process and switching between the dilution branch and the long or short optical path flow cell; performing formal acquisition after preheating, pre-stabilization, residual release, and sample replacement volume reaching a preset multiple; calculating the concentration of the analyte based on the repeatable spectral characteristic response and corresponding calibration parameters, and performing repeatability verification; determining the validity of the concentration result by combining each state level with the verification result; and outputting retest, anomaly handling, or cleaning results if the validity conditions are not met.

Description

Technical Field

[0001] This invention relates to the field of agricultural fertilizer detection and spectral analysis technology, specifically a method for detecting the components of high-potassium organic water-soluble fertilizer based on spectral analysis. Background Technology

[0002] With the increasing application of high-potassium organic water-soluble fertilizers in fertilizer production and quality control, component detection in the production process is gradually developing towards real-time online detection. These samples typically have characteristics such as high ionic strength, high concentration, and deep color. When using spectral analysis for detection, how to ensure detection efficiency while also ensuring the accuracy of results and the stability of equipment operation is a technical problem that urgently needs to be solved by those skilled in the art. The current spectral detection of high-potassium organic water-soluble fertilizers mainly relies on the following methods: after the sample enters the detection channel, a fixed optical path measurement is performed directly; when the signal is abnormal, manual or preset dilution and retesting are performed; after the detection is completed, the sample is rinsed and a simple blank is confirmed before proceeding to the next sample for detection. However, methods such as fixed optical path direct measurement, post-detection dilution retesting, and routine rinsing and release all have certain drawbacks. For example, the fixed optical path method is prone to problems such as excessively high absorbance, low transmittance, and signal saturation when dealing with high-concentration, dark-colored samples. The post-detection dilution retesting method usually lacks joint judgment on process states such as temperature, pressure difference, bubbles, and reference light intensity, which can easily introduce repeatability deviations when the flow is unstable or the sample replacement is insufficient. The routine rinsing and simple blank confirmation methods are difficult to effectively identify the risk of cross-contamination caused by residues from the previous sample, wall adhesion, or crystallization, thus affecting the reliability of subsequent sample detection results. Summary of the Invention

[0003] The purpose of this invention is to provide a method for detecting high-potassium organic water-soluble fertilizer components based on spectral analysis, thereby solving the technical problems of existing technologies, such as absorption saturation when dealing with high-concentration, dark-colored samples, and repeatability bias and cross-contamination due to the lack of joint judgment of process states. To this end, the technical solution of this invention includes: A method for detecting high-potassium organic water-soluble fertilizer components based on spectral analysis is implemented on a detection device equipped with a sample inlet path, a dilution branch path, a long-path flow cell, a short-path flow cell, a main detection optical path, a reference optical path, a cleaning path, and units for acquiring conductivity, temperature, pressure difference, and bubble signals. The method includes: The conductivity, temperature, pressure difference before and after the long-path or short-path flow cell, bubble passing signal, reference light intensity, blank background spectrum and sample spectrum are collected during the sample detection process. Based on the conductivity, temperature, pressure difference across the long-path or short-path flow cell, bubble passage signal, reference light intensity, blank background spectrum, and sample spectrum, the transmittance load level, flow stability, and residual contamination level are determined respectively. Based on the light transmittance level, flow stability, and residual contamination level, select the corresponding detection process and switch the dilution branch and the long-path flow cell or the short-path flow cell; after meeting the preset preheating conditions, preset flow stabilization conditions, preset residual release conditions, and the sample replacement volume reaches the preset multiple, perform multiple formal collections on the same sample. Based on the repeated spectral characteristic responses of the same sample obtained from multiple formal acquisitions, and according to the pre-stored calibration parameters corresponding to the current optical path and current dilution ratio, the concentration of the analyte is calculated and repeatability is verified. The validity of the concentration results is determined based on the light transmittance level, flow stability, residual contamination level, and repeatability test results. If the concentration result is invalid, output the result of retesting, anomaly handling, or cleaning.

[0004] Preferably, the detection device further includes a premixing unit and a pre-stabilizing cavity; the steps of acquiring conductivity, temperature, pressure difference before and after the long-path or short-path flow cell, bubble passage signal, reference light intensity, blank background spectrum and sample spectrum during the sample detection process include: acquiring conductivity and temperature after the sample enters the premixing unit; acquiring pressure difference before and after the long-path or short-path flow cell and bubble passage signal before and after the sample enters the pre-stabilizing cavity and the long-path or short-path flow cell; acquiring reference light intensity, blank background spectrum and sample spectrum before and during spectral acquisition; and storing the above acquisition results corresponding to the same detection sample.

[0005] Preferably, the light transmittance level is determined based on the average transmittance or average absorbance of the preset short-wavelength characteristic region of the sample spectrum, the average light intensity or average absorbance of the preset background band, and the light intensity variation between the main detection optical path and the reference optical path. By combining the first parameter characterizing the average transmittance or average absorbance of the preset short-wavelength characteristic region, the second parameter characterizing the average light intensity or average absorbance of the preset background band, and the third parameter characterizing the relative light intensity change between the main detection optical path and the reference optical path, the transmittance load level is determined to be low load, medium load, or high load.

[0006] Preferably, the flow stability is determined based on the pressure difference fluctuations, temperature fluctuations, bubble passing frequency, and reference light intensity fluctuations across the long-path or short-path flow cell within a preset observation time window. The flow stability is determined to be steady-state, disturbed state, or transitional state by comprehensively considering the pressure difference fluctuations, temperature fluctuations, bubble passing frequency, and reference light intensity fluctuations across the long-path or short-path flow cell. When the bubble passing frequency continuously exceeds a preset bubble frequency threshold, or when the pressure difference across the long-path or short-path flow cell continuously exceeds a preset pressure difference blockage threshold, the flow stability is determined to be disturbed state.

[0007] Preferably, the residual contamination level is determined based on the blank background deviation after cleaning, the rinsing recovery deviation, and the contamination parameters of the previous sample; The contamination parameters of the previous sample are parameters that characterize the residual risk of the previous sample to be tested, including at least one of the following: the peak absorbance of the previous sample in the preset contamination characteristic band, the conductivity of the previous sample, the viscosity grade of the previous sample, and the change in pressure difference of the flow cell before and after cleaning. Based on the combined background deviation, rinsing recovery deviation, and contamination parameters of the previous sample, the residual contamination level is determined to be low risk, medium risk, or high risk.

[0008] Preferably, the preset preheating condition is: the fluctuation of the reference light intensity within the preset preheating window is not greater than the preset preheating limit; The preset current stabilization conditions are as follows: the deviation between the average pressure difference across the long-path or short-path flow cell within the preset observation time window and the target steady-state pressure difference is not greater than the pressure difference pre-stabilization limit; the deviation between the average temperature within the preset observation time window and the target temperature control temperature is not greater than the temperature pre-stabilization limit; and the bubble passage frequency is not greater than the bubble frequency pre-stabilization limit, wherein the deviation is a difference or an absolute difference. The preset residual release conditions are as follows: the blank background deviation and the rinsing recovery deviation do not exceed the corresponding release limits. Only when the preset preheating condition, the preset current stabilization condition, and the preset residual release condition are all met simultaneously, it is determined that the prerequisites for performing formal data acquisition are met.

[0009] Preferably, the detection process includes a rapid detection process, a verification detection process, and an anomaly handling process; When the light transmittance load is low or medium, the flow stability is steady, and the residual contamination level is low risk, choose the rapid detection process. When the light transmittance level is high or the residual contamination level is medium or high risk, select the verification testing process; When the flow is in a stable state but is in a transitional or disturbed state, select the abnormal handling procedure; When the absorbance of the sample spectrum in the preset short-wavelength characteristic region under the long-path flow cell exceeds the preset upper limit, the sample is switched to the short-path flow cell. When the absorbance of the sample in the preset short-wavelength characteristic region under the short-path flow cell reaches the upper limit of the instrument calibration or the corresponding transmittance is lower than the lower limit of the instrument calibration, and the absorption saturation state is determined to have occurred, a dilution and retest is performed.

[0010] Preferably, before performing formal collection, the sample is first introduced into the pre-stabilized flow chamber for buffering and flow rate adjustment for a preset time; Once the preset current stabilization conditions are met and maintained for a preset time, the sample is switched to the formal measurement channel consisting of a long optical path flow cell or a short optical path flow cell. After the sample enters the long-path flow cell or the short-path flow cell, it first undergoes a circulation flushing and replacement process to make the replacement volume reach a preset multiple of the effective volume of the flow cell and connecting pipeline before formal integration and acquisition are started.

[0011] Preferably, the conversion of the concentration of the analyte includes: The corrected characteristic response is obtained based on the characteristic peak absorbance of the analyte and the absorbance at the background wavelength. The concentration of the analyte and the repeatability verification results are obtained according to the corresponding pre-stored calibration parameters and the current dilution ratio.

[0012] Preferably, when the concentration result is invalid, a cleaning and recovery step is performed, including: The cleaning level is determined based on the residual contamination level, the contamination parameters of the previous sample, whether crystallization or wall adhesion occurs, and the blank background deviation and rinsing recovery deviation after cleaning. When the cleaning level is low, perform a single round of pure water rinsing and a blank confirmation. When the cleaning level is medium, perform pure water rinsing and low-speed replacement, and perform two blank checks. When the cleaning level is high, perform enhanced cleaning and blank verification. The wall contamination deviation is calculated by extracting the spectral difference of the wall contamination sensitive band. The wall contamination sensitive band refers to the preset characteristic band that is affected by water-soluble fertilizer deposits and whose absorbance changes exceed the preset deviation threshold. The cleaning result is determined based on the blank background deviation, the rinsing recovery deviation and the wall contamination deviation. When the cleaning result meets the release requirements, the sample injection path is opened.

[0013] Compared with the prior art, the present invention has the following beneficial effects: The method for detecting high-potassium organic water-soluble fertilizer components based on spectral analysis in this application involves jointly collecting conductivity, temperature, pressure difference, bubble transmission signal, reference light intensity, blank background spectrum and sample spectrum before formal collection, and determining the light transmittance load level, flow stability and residual pollution level accordingly. Further linkage switching of dilution branch and long-path flow cell or short-path flow cell, selection of rapid detection process, verification detection process or abnormal handling process, and formal acquisition is only performed after preheating, pre-stabilization, residual release and sample replacement volume reach the preset multiple. Combined with repeatable spectral characteristic response, corresponding calibration parameters, repeatability verification and graded cleaning and recovery mechanism to complete result release or retesting, abnormal handling and cleaning control. This effectively avoids problems such as excessively high absorbance, low transmittance, and signal saturation that occur when directly fixing the optical path for high-concentration dark samples. It also reduces the risk of repeatability deviations and cross-contamination caused by flow instability, bubble disturbance, temperature fluctuation, insufficient replacement, and residues from the previous sample, wall adhesion, or crystallization. This improves the accuracy, stability, and continuous operation capability of high-potassium organic water-soluble fertilizer component detection. Attached Figure Description

[0014] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. The drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0015] Figure 1 This is a schematic flowchart of the method for detecting high-potassium organic water-soluble fertilizer components based on spectral analysis provided in an embodiment of the present invention. Detailed Implementation

[0016] To make the objectives, technical solutions, and advantages of the present invention clearer, the technical solutions of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0017] Please see the appendix Figure 1 A method for detecting high-potassium organic water-soluble fertilizer components based on spectral analysis is implemented on a detection device equipped with a sample inlet flow path, a dilution branch, a long-path flow cell, a short-path flow cell, a main detection optical path, a reference optical path, a cleaning flow path, and units for acquiring conductivity, temperature, pressure difference, and bubble signals. The method includes: The conductivity, temperature, pressure difference before and after the long-path or short-path flow cell, bubble passing signal, reference light intensity, blank background spectrum and sample spectrum are collected during the sample detection process. Based on conductivity, temperature, pressure difference across the long-path or short-path flow cell, bubble passage signal, reference light intensity, blank background spectrum, and sample spectrum, the transmittance load level, flow stability, and residual contamination level are determined respectively. Based on the light transmittance level, flow stability, and residual contamination level, select the corresponding detection process and switch the dilution branch and the long-path flow cell or the short-path flow cell; after meeting the preset preheating conditions, preset flow stabilization conditions, preset residual release conditions, and the sample replacement volume reaches the preset multiple, perform multiple formal collections on the same sample. Based on the repeated spectral characteristic responses of the same sample obtained from multiple formal acquisitions, and according to the pre-stored calibration parameters corresponding to the current optical path and current dilution ratio, the concentration of the analyte is calculated and repeatability is verified. The validity of the concentration results is determined based on the light transmittance level, flow stability, residual contamination level, and repeatability test results. When the concentration result is invalid, the system outputs the result for retesting, anomaly handling, or cleaning; the detection equipment also includes a premixing unit and a pre-stabilizing chamber. The steps for collecting conductivity, temperature, pressure difference before and after the long-path or short-path flow cell, bubble passage signal, reference light intensity, blank background spectrum, and sample spectrum during the sample detection process include: collecting conductivity and temperature after the sample enters the premixing unit; collecting pressure difference before and after the long-path or short-path flow cell and bubble passage signal before and after the sample enters the pre-stabilized flow chamber and before and after the long-path or short-path flow cell; collecting reference light intensity, blank background spectrum, and sample spectrum before and during spectral acquisition; and storing the above acquisition results according to the same detection sample. The transmittance level is determined based on the average transmittance or average absorbance of the preset short-wavelength characteristic region of the sample spectrum, the average light intensity or average absorbance of the preset background band, and the light intensity change between the main detection optical path and the reference optical path. The transmittance level is determined to be low, medium, or high by combining the first parameter characterizing the average transmittance or average absorbance of the preset short-wavelength characteristic region, the second parameter characterizing the average light intensity or average absorbance of the preset background band, and the third parameter characterizing the relative light intensity change between the main detection optical path and the reference optical path. In the method for detecting components of high-potassium organic water-soluble fertilizer based on spectral analysis, the transmittance of the sample is evaluated and the detection path is dynamically adapted before formal acquisition by extracting signals from multiple sensors. The specific process includes: The system collects a series of fluid and optical state data on a detection device equipped with hardware modules such as a sample inlet flow path, a dilution branch, a long-path flow cell, and a short-path flow cell. After the sample is pumped into the premixing unit, the system extracts its conductivity and temperature. Before and after the sample flows through the pre-stabilization cavity and the long and short-path flow cells, the system continuously acquires the pressure difference and bubble passage signals. During the spectral acquisition phase, the system simultaneously acquires the reference light intensity, blank background spectrum, and sample spectrum. These multi-source physical quantities are uniformly stored and associated with the current detection task. Based on the aforementioned multidimensional data, the system calculates the light transmittance level, flow stability, and residual contamination level according to the rules. Specifically, for the light transmittance level, the system extracts the transmission information of the preset short-wavelength characteristic region of the sample spectrum, the preset background band information, and the light intensity changes between the main detection optical path and the reference optical path. The system unifies the dimensions of the first, second, and third parameters characterizing the above features according to their respective preset upper and lower limits, and calculates the part exceeding the limit value. This unified dimensional processing includes alignment mapping of positive and negative indices; the system maps all parameters to a dimensionless load score range of 0-100; its general linear mapping formula for dimensional processing is: in, The dimensionless score after mapping. The current measured value of the parameter. and These are the preset lower limit and preset upper limit for the parameter, respectively. For example, if the first parameter is selected as the average absorbance of the preset short-wave characteristic region, which is positively correlated with the light transmittance, the system uses the above positive correlation formula to perform linear conversion, mapping its preset lower limit to 0 and its preset upper limit to 100. If the average transmittance is selected, it is negatively correlated, and the mapping rule is reversed. The system uses the following negative correlation linear mapping formula for conversion. A single transmittance value is formed by weighted summation or weighted averaging; assuming the average absorbance of a high-potassium sample in the short-wavelength characteristic region is 1.8, and preset upper and lower limits are set to 0.5 and 2.0, its dimensionally processed mapping value is: Assuming the mapping value of the second parameter is 70.0 and the mapping value of the third parameter is 80.0; the weight allocation of each parameter in the weighted summation is based on the following: the first parameter reflects the transmittance of the main absorption characteristic region of the analyte and is directly related to the spectral signal-to-noise ratio, so it is assigned the highest weight; the second parameter reflects the light scattering or overall attenuation of the background across the entire spectral band, so it is assigned a secondary weight. The third parameter reflects the transient deviation of the instrument's own optical path system and is assigned the lowest weight. Based on the aforementioned weight allocation criteria, if the weights of these three parameters are 0.5, 0.3, and 0.2 respectively, then the calculated single transmittance load value is: The transmittance value is compared sequentially with the first load limit and the second load limit. The preset first and second load limits are determined through preliminary calibration experiments: by conducting a series of tests on a range of dark-colored fertilizer standard samples with known gradient concentrations, the critical transmittance value when the absorbance in the short-wavelength characteristic region under the long optical path flow cell reaches 80% of the instrument's linear detection limit is recorded, and this value is set as the first load limit; the transmittance value when the absorbance is at the saturation critical point is recorded, and this value is set as the second load limit; if the preset first load limit is 30 and the second load limit is 75, since 80.35 is greater than 75, the transmittance state of the current sample is classified as high load. Based on the above three dimensions of state levels, the system selects the corresponding detection process and instructs the relevant valves to switch the dilution branch and the long or short optical path flow cell; after confirming that the preheating, pre-stabilization, and residual release conditions are all met, and the displacement volume of the sample in the pipeline reaches the set multiple, the system triggers the formal spectral acquisition action. After acquisition, the system calls the pre-stored calibration parameters that match the current optical path and dilution ratio based on the repeatable spectral characteristic response of the same sample, calculates the concentration of the analyte, and performs repeatability verification. Finally, it combines the three previous status levels and the repeatability verification results to determine whether the current concentration result is valid. If invalid, the system stops outputting concentration data and instead issues retest, anomaly handling, or cleaning instructions. To ensure the continuity of the evaluation process, if data from a sensor such as a temperature sensor is momentarily lost during the acquisition process, the system will pause the current state determination, call the stable value from the previous valid sampling period as a replacement, and record the data calculation identifier; if the light transmittance load parameter continuously exceeds the upper limit of dimensional processing and is in a continuous overflow state, the system will directly terminate the current detection task and trigger pipeline flushing. In a specific application scenario: on a semi-online testing production line at a high-potassium organic water-soluble fertilizer production site, the test solution typically has high ionic strength and a dark color. When a batch of samples is pumped into the testing equipment, in order to eliminate the problem of unobservable temporal state caused by the lack of prior state acquisition, the system does not immediately start a high-precision long-term formal spectral scan. Instead, it first controls the spectrometer to perform a rapid pre-irradiation of the solution with an integration time shorter than the formal acquisition integration time, and extracts the initial sample spectrum for state assessment. Meanwhile, the basic conductivity is obtained in the premixing unit, and the pressure difference and bubbles are monitored in the flow channel. Based on the pre-scanned optical and fluid data, the system calculates that the sample is in a high-load state. Then, the system schedules the dilution branch to introduce diluent and switches to the short optical path flow cell. Only after all fluid and optical conditions meet the standards is the formal spectral scanning command for concentration conversion issued. This avoids spectral saturation caused by direct measurement of dark samples in the physical control link.

[0018] The flow stability is determined based on the pressure difference fluctuations, temperature fluctuations, bubble passing frequency, and reference light intensity fluctuations across the long-path or short-path flow cell within a preset observation time window. The flow stability is determined as steady-state, disturbed state, or transitional state by comprehensively considering these factors. When the bubble passing frequency continuously exceeds a preset bubble frequency threshold, or when the pressure difference across the long-path or short-path flow cell continuously exceeds a preset pressure difference clogging threshold, the flow stability is determined to be disturbed state. The residual contamination level is determined based on the blank background deviation after cleaning, the rinsing recovery deviation, and the contamination parameters of the previous sample. The previous sample contamination parameters are parameters that characterize the residual risk of the previous sample to be tested, including at least one of the following: the peak absorbance of the previous sample in the preset contamination characteristic band, the conductivity of the previous sample, the viscosity grade of the previous sample, and the change in pressure difference of the flow cell before and after cleaning. Based on the combined background deviation, rinsing recovery deviation, and contamination parameters of the previous sample, the residual contamination level is determined to be low risk, medium risk, or high risk; the preset preheating conditions are: the reference light intensity fluctuation within the preset preheating window does not exceed the preset preheating limit; The preset stabilization conditions are: the deviation between the average pressure difference across the long-path or short-path flow cell within the preset observation time window and the target steady-state pressure difference is not greater than the pressure difference pre-stabilization limit; the deviation between the average temperature within the preset observation time window and the target temperature control temperature is not greater than the temperature pre-stabilization limit; and the bubble passage frequency is not greater than the bubble frequency pre-stabilization limit, where the deviation is a difference or an absolute difference. The preset residual release conditions are: the blank background deviation and the rinsing recovery deviation do not exceed the corresponding release limits. Only when the preset preheating conditions, preset stabilization conditions, and preset residual release conditions are all met simultaneously, it is determined that the prerequisites for performing formal data acquisition are met. The testing process includes a rapid testing process, a verification testing process, and an anomaly handling process. When the light transmittance level is low or medium, the flow stability is steady-state, and the residual contamination level is low risk, the rapid testing process is selected. When the light transmittance level is high or the residual contamination level is medium or high risk, the verification testing process is selected. When the flow stability is transitional or disturbed, the anomaly handling process is selected. When the absorbance of the sample spectrum in the preset short-wavelength characteristic region under the long optical path flow cell exceeds the preset upper limit, switch to the short optical path flow cell; when the absorbance of the sample in the preset short-wavelength characteristic region under the short optical path flow cell reaches the upper limit of the instrument calibration or the corresponding transmittance is lower than the lower limit of the instrument calibration, and the absorption saturation state is determined to have occurred, perform dilution and retest. To avoid microbubbles or localized crystallization forming when samples flow through pipelines experience pressure drops, reversals, or temperature fluctuations, which could introduce fluid disturbances into the spectral signal, the following logic governs the assessment of flow stability and residual contamination levels, as well as the coordinated scheduling of the detection process: Within a set observation time window, the system extracts the range, standard deviation, or coefficient of variation of pressure difference, temperature, and reference light intensity, and counts the frequency of bubble passage per unit time. These four types of fluctuation indicators are weighted after being processed with unified dimensions to form a flow stability value. In this processing, the range, standard deviation, coefficient of variation, and bubble passage frequency are all positive disturbance indicators, that is, the larger the indicator value, the more violent the fluid fluctuation. The system pre-calibrates the normal fluctuation lower limit (mapped to 0 points) and the tolerable disturbance upper limit (mapped to 100 points) based on experiments. It then converts the fluctuation indicators extracted from each channel into a disturbance score range of 0-100 according to a linear ratio. Values ​​exceeding the upper limit are directly counted as 100. The weighting allocation is determined based on the degree of interference of each indicator on the spectral signal-to-noise ratio: pressure difference fluctuation directly characterizes the overall pulsation and flow velocity unevenness of the fluid, and has the greatest impact on absorbance baseline drift, so it is assigned the highest weight; the frequency of bubble passage causes transient truncation of optical path, generating severe noise, so it is assigned the second highest weight; temperature fluctuation and reference light intensity fluctuation are assigned relatively small weights. For example, within a 5-second observation window, if the coefficient of variation mapping score for pressure fluctuation is 20, the temperature fluctuation range mapping score is 15, the bubble passage frequency mapping score is 0, and the reference light intensity fluctuation standard deviation mapping score is 10, and based on the aforementioned allocation criteria, if the weights of these four items are set to 0.4, 0.2, 0.3, and 0.1 respectively, then the weighted summation of the flow stability value is... The system compares the stable value with the first and second stability limits, and, in conjunction with the duration condition, marks the fluid state as steady state, transition state, or disturbance state; wherein, the first and second stability limits are calibrated based on the baseline fluctuation distribution measured by the equipment under deionized water pure operating conditions and in conjunction with the allowable maximum spectral relative standard deviation. Assuming the preset first stability threshold is 15 and the second stability threshold is 60, since the current value 12 is not greater than the first stability threshold 15, the system monitors that it remains below this threshold for, for example, three consecutive observation time windows for a first preset time, thus confirming that the fluid has entered a steady state. At the same time, the system extracts the difference between the blank background spectrum after cleaning and the standard blank spectrum in the sensitive band to calculate the blank background deviation, calculates the rinsing recovery deviation through the conductivity of the effluent, and reads historical residual risk parameters of the previous sample, such as absorbance peak value and viscosity grade. These three parameters are weighted according to a predetermined weight set based on the correlation coefficient of historical residual detection of each indicator to form a residual contamination value. This unified dimension processing also adopts a linear mapping of dimensionless scores from 0 to 100 based on preset boundaries; for example, the blank background deviation mapping score is calculated based on the ratio of its difference to the maximum allowable difference, and if it exceeds the maximum allowable difference, it is recorded as 100; the rinsing recovery deviation is similarly mapped based on the absolute difference in conductivity between the effluent and the standard blank solution. For the historical residual risk parameters of the previous sample, if the absorbance peak exceeds 90% of the instrument's maximum range, then this item is directly mapped to 100. Let the mapped scores of the aforementioned three quantified items be 15, 25, and 80, respectively, and the system's preset weights be 0.3, 0.3, and 0.4, respectively. Then the weighted calculated residual contamination value is... The preset first and second risk limits are obtained by calibrating the blank background baseline deviation from the allowable range caused by the introduction of known gradient concentrations of pollutants into the standard blank solution. The residual pollution value is divided into low, medium and high risk levels. Before the formal collection is to be carried out, the system strictly verifies the release conditions: the reference light intensity fluctuation, the deviation between the average pressure difference and the target pressure difference, the deviation between the average temperature and the target temperature, the bubble frequency, and the aforementioned blank and rinsing deviations within the preheating window are extracted. The system generates a data acquisition permission instruction only when all of the above indicators are not greater than their respective pre-stabilization or release limits. In terms of detection process scheduling, the system performs flow diversion based on the above status labels: when the light transmittance is low or medium, the fluid is in a steady state, and the residue is not high-risk, the system executes a rapid detection process. If the light transmittance is high, or the residual risk reaches a medium or high level, the system controls the execution of the verification and testing process; once the fluid is determined to be a disturbance, the abnormal handling process is forcibly switched; in the verification process, if the absorbance of the short-wavelength characteristic region exceeds the limit under the long optical path, the system issues an instruction to switch to the short optical path; if the short optical path still results in absorbance saturation, the system further calls the dilution branch to perform dilution retest. In addition, to avoid control link conflicts between long-term evaluation during the observation window and transient equipment damage, the system establishes a dual-layer collaborative monitoring mechanism for bubble and differential pressure data. If, during the observation window under stable flow conditions, the control unit detects that the frequency of bubble passage exceeds the preset threshold instantaneously or the differential pressure exceeds the limit blockage threshold instantaneously, the system will no longer wait for the main control logic judgment cycle to end. Instead, it will directly interrupt the current judgment process through hardware interruption, forcibly mark the state as a disturbance, and immediately issue an emergency stop command for drainage to avoid the risk of communication delay and mismatch between control actions. If the system finds that the historical parameters of the previous sample are missing when calculating the residual contamination value, it will handle it as a high-risk level and trigger the pipeline protection program. In a specific application scenario: on the high-potassium organic water-soluble fertilizer testing production line, the previous batch tested dark-colored samples with high viscosity and high humic acid content; when the current batch of low-concentration samples enters the equipment, the system reads the high-risk parameters of the previous batch, calculates that the current pipeline residual pollution value is high-risk, and thus refuses to execute the rapid testing process and forces the process to enter the verification process. Meanwhile, during the pre-stabilization period, the system detected that the rate of change of differential pressure in the pipeline exceeded the preset threshold and the frequency of bubbles exceeded the limit. It was determined that the fluid was in a disturbed state. The system then paused the spectral acquisition and started the abnormal handling process to degas and circulate the pipeline until all deviation indicators fell back to within the release limit before subsequent concentration measurement was allowed.

[0019] Before formal acquisition, the sample is first introduced into the pre-stabilized flow chamber for buffering and flow rate adjustment for a preset time. After the preset stabilization conditions are met and maintained for a preset time, the sample is switched to the formal measurement channel consisting of a long optical path flow cell or a short optical path flow cell. After the sample enters the long optical path flow cell or the short optical path flow cell, it is first circulated and flushed to make the replacement volume reach a preset multiple of the effective volume of the flow cell and connecting pipeline before formal integration acquisition is started. The conversion of the concentration of the analyte includes: obtaining the corrected characteristic response based on the characteristic peak absorbance of the analyte and the absorbance at the background wavelength point, and obtaining the concentration of the analyte and the repeatability verification result according to the corresponding pre-stored calibration parameters and the current dilution ratio; When the concentration result is invalid, a cleaning and recovery procedure is performed, including: determining the cleaning level based on the residual contamination level, the contamination parameters of the previous sample, whether crystallization or wall adhesion occurs, and the blank background deviation and rinsing recovery deviation after cleaning; when the cleaning level is low, a single round of pure water rinsing is performed and a blank confirmation is performed; when the cleaning level is medium, pure water rinsing and low-speed displacement are performed and two blank confirmations are performed; when the cleaning level is high, enhanced cleaning and blank confirmation are performed. The wall contamination sensitive band refers to the preset characteristic band that is affected by water-soluble fertilizer deposits and whose absorbance changes exceed the preset deviation threshold; the spectral difference of the wall contamination sensitive band is extracted to calculate the wall contamination deviation, and the cleaning result is determined based on the blank background deviation, rinsing recovery deviation and wall contamination deviation. When the cleaning result meets the release requirements, the sample injection path is opened. After completing the condition assessment and path allocation, to avoid concentration boundary layer interference caused by the mixed liquid at the pipeline front and cross-contamination of samples after testing and release, the following steady-flow replacement, concentration conversion verification, and graded cleaning steps are performed: Before the formal collection, the system control valve introduces the sample into the pre-stabilized flow chamber for flow stabilization. Once the pre-stabilized flow condition is consistently met, the system switches the flow path to the formal measurement channel, which consists of a long optical path or a short optical path flow cell. At this point, the system does not immediately start the spectrometer. Instead, it instructs the pump group to perform cyclic flushing and replacement until the volume of the replaced sample reaches a preset multiple of the effective volume of the flow cell and tubing. This preset multiple is determined based on the Reynolds number and fluid staining experiments of the tubing dead volume distribution, and is generally taken as 3 to 5 times to ensure that the laminar flow adhesion effect is completely eliminated and the concentration boundary layer reaches stability, triggering formal integration and acquisition. After acquiring the spectrum, the system extracts the characteristic peak absorbance of the target component and the absorbance of the background wavelength point for difference correction to eliminate baseline drift. The system calls the pre-stored calibration parameters that match the current optical path and dilution ratio, and converts the corrected characteristic response into the target component concentration; if the dilution branch is enabled, it is further multiplied by the total dilution ratio to restore the original concentration; the system calculates the repeatability deviation of the multiple conversion results of the same sample under the same conditions and compares it with the evaluation limit. When a concentration result is deemed invalid due to exceeding repeatability limits or failing to meet preconditions, the system freezes the result output and proceeds with cleaning and recovery. The system determines the cleaning level based on the residual contamination level, the contamination parameters of the previous sample, whether crystallization has occurred, and the blank recovery status. The specific judgment logic is as follows: First, it is determined whether crystallization has occurred or whether the contamination parameters of the previous sample indicate high viscosity. If crystallization is determined due to an abnormal increase in the preceding pressure difference, or if the viscosity level reaches the preset viscosity threshold, then it is directly locked as a high-level cleaning. If not, further determine whether the residual contamination level is high-risk or medium-risk, or whether the initial flushing recovery deviation exceeds the warning limit. If either condition is met, it is determined to be medium-level cleaning. If none of the above conditions are triggered, and the residual contamination level is low-risk, it is determined to be low-level cleaning. Based on the above classification, low-level cleaning triggers a single round of pure water flushing and one blank confirmation. Medium-level flushing triggers pure water rinsing, issues control commands to reduce the flow rate to 30%-50% of the normal injection flow rate to extend the wetting time, and performs low-speed replacement and two blank confirmations. High-level flushing activates enhanced cleaning resources, specifically including issuing quantitative control commands to increase the pump speed to 80%-100% of the rated maximum speed, introduce hot cleaning fluid to perform pulse rinsing such as 3-5 preset times, and extend the single rinsing time to at least 3 times the low-level cleaning time. During enhanced cleaning, in order to ensure the structural safety of the optical equipment while avoiding a sudden increase in local resistance in the flow channel, the system simultaneously issues a high-speed pulse flushing command and simultaneously opens the proportional pressure relief bypass connected in parallel with the flow cell according to the preset resistance curve to ensure that the transient fluid impact pressure difference at both ends of the flow cell is always limited to below the safety stress threshold of the quartz window and sealing ring. After cleaning, the system extracts the spectral difference of the wall contamination sensitive band to calculate the wall contamination deviation, and evaluates the cleaning recovery result by combining the blank background deviation and the rinsing recovery deviation; only when the result meets the release criteria will the system release the flow path occupancy state and allow the next sample to be tested to enter; If the system detects an abnormal increase in pressure difference during the cyclic flushing and replacement process, it indicates that there may be pipe adhesion or crystallization. The system will stop the replacement process and directly jump to the higher-level enhanced cleaning step. If the cleaning recovery results still cannot meet the release requirements after multiple rounds of cleaning, the system will lock the sample inlet channel and send a signal to the host computer for manual intervention to block new samples from entering the flow path.

[0020] In a specific application scenario: On a high-potassium organic water-soluble fertilizer production line, a high-concentration sample is introduced into a pre-stabilized flow chamber to eliminate pumping pulsation and microbubbles; after switching to the measurement channel, the system pumps in a sample of a preset multiple of the pipeline volume to flush away the dilution mixture at the leading edge; after the spectral acquisition is completed, the system extracts the true potassium salt characteristic response by subtracting the background absorbance and calculates the concentration by combining the calibration parameters. Because the sample concentration exceeded the preset range limit and was accompanied by local crystallization, the repeatability verification failed. The system determined the result to be invalid and terminated the data upload. The system automatically classified it as a high-level cleaning and called the cleaning solution to strengthen the wetting of the flow cell and replace it with pure water. Only when the spectral scan confirmed that the deposits on the inner wall of the flow cell had been removed and the wall contamination deviation met the standard, did the system open the entrance and allow the next batch of fertilizer samples to enter for testing.

[0021] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for detecting components of high-potassium organic water-soluble fertilizer based on spectral analysis, characterized in that, Implemented on a detection device comprising a sample inlet path, a dilution branch, a long-path flow cell, a short-path flow cell, a main detection optical path, a reference optical path, a cleaning path, and units for acquiring conductivity, temperature, pressure difference, and bubble signals, including: The conductivity, temperature, pressure difference before and after the long-path or short-path flow cell, bubble passing signal, reference light intensity, blank background spectrum and sample spectrum are collected during the sample detection process. Based on the conductivity, temperature, pressure difference across the long-path or short-path flow cell, bubble passage signal, reference light intensity, blank background spectrum, and sample spectrum, the transmittance load level, flow stability, and residual contamination level are determined respectively. Based on the light transmittance level, flow stability, and residual contamination level, select the corresponding detection process and switch the dilution branch and the long-path flow cell or the short-path flow cell; after meeting the preset preheating conditions, preset flow stabilization conditions, preset residual release conditions, and the sample replacement volume reaches the preset multiple, perform multiple formal collections on the same sample. Based on the repeated spectral characteristic responses of the same sample obtained from multiple formal acquisitions, and according to the pre-stored calibration parameters corresponding to the current optical path and current dilution ratio, the concentration of the analyte is calculated and repeatability is verified. The validity of the concentration results is determined based on the light transmittance level, flow stability, residual contamination level, and repeatability test results. If the concentration result is invalid, output the result of retesting, anomaly handling, or cleaning.

2. The method according to claim 1, characterized in that... The detection device further includes a premixing unit and a pre-stabilizing cavity; the steps of collecting conductivity, temperature, pressure difference before and after the long-path or short-path flow cell, bubble passage signal, reference light intensity, blank background spectrum and sample spectrum during the sample detection process include: collecting conductivity and temperature after the sample enters the premixing unit; collecting pressure difference before and after the long-path or short-path flow cell and bubble passage signal before and after the sample enters the pre-stabilizing cavity and the long-path or short-path flow cell; collecting reference light intensity, blank background spectrum and sample spectrum before and during spectral acquisition; and storing the above acquisition results according to the same detection sample.

3. The method according to claim 1, characterized in that, The transmittance level is determined based on the average transmittance or average absorbance of the preset short-wavelength characteristic region of the sample spectrum, the average light intensity or average absorbance of the preset background band, and the light intensity variation between the main detection optical path and the reference optical path. By combining the first parameter characterizing the average transmittance or average absorbance of the preset short-wavelength characteristic region, the second parameter characterizing the average light intensity or average absorbance of the preset background band, and the third parameter characterizing the relative light intensity change between the main detection optical path and the reference optical path, the transmittance load level is determined to be low load, medium load, or high load.

4. The method according to claim 1, characterized in that, The flow stability is determined based on the pressure difference fluctuations, temperature fluctuations, bubble passing frequency, and reference light intensity fluctuations across the long-path or short-path flow cell within a preset observation time window. The flow stability is determined as steady-state, disturbed state, or transitional state by comprehensively considering these factors. When the bubble passing frequency continuously exceeds a preset bubble frequency threshold, or when the pressure difference across the long-path or short-path flow cell continuously exceeds a preset pressure difference blockage threshold, the flow stability is determined to be disturbed state.

5. The method according to claim 1, characterized in that, The residual contamination level is determined based on the blank background deviation after cleaning, the rinsing recovery deviation, and the contamination parameters of the previous sample. The contamination parameters of the previous sample are parameters that characterize the residual risk of the previous sample to be tested, including at least one of the following: the peak absorbance of the previous sample in the preset contamination characteristic band, the conductivity of the previous sample, the viscosity grade of the previous sample, and the change in pressure difference of the flow cell before and after cleaning. Based on the combined background deviation, rinsing recovery deviation, and contamination parameters of the previous sample, the residual contamination level is determined to be low risk, medium risk, or high risk.

6. The method according to claim 5, characterized in that, The preset preheating condition is: the fluctuation of the reference light intensity within the preset preheating window is not greater than the preset preheating limit. The preset current stabilization conditions are: the deviation between the average pressure difference across the long-path flow cell or short-path flow cell within the preset observation time window and the target steady-state pressure difference is not greater than the pressure difference pre-stabilization limit; the deviation between the average temperature within the preset observation time window and the target temperature control temperature is not greater than the temperature pre-stabilization limit; and the frequency of bubble passage is not greater than the bubble frequency pre-stabilization limit, wherein the deviation is a difference or an absolute difference. The preset residual release conditions are: the blank background deviation and the rinsing recovery deviation do not exceed the corresponding release limits; the prerequisite for performing formal data acquisition is determined to be met only when the preset preheating condition, the preset flow stabilization condition, and the preset residual release condition are met simultaneously.

7. The method according to claim 1, characterized in that, The detection process includes a rapid detection process, a verification detection process, and an anomaly handling process. When the light transmittance load is low or medium, the flow stability is steady, and the residual contamination level is low risk, choose the rapid detection process. When the light transmittance level is high or the residual contamination level is medium or high risk, select the verification testing process; When the flow is in a stable state but is in a transitional or disturbed state, select the abnormal handling procedure; When the absorbance of the sample spectrum in the preset short-wavelength characteristic region under the long-path flow cell exceeds the preset upper limit, the sample is switched to the short-path flow cell. When the absorbance of the sample in the preset short-wavelength characteristic region under the short-path flow cell reaches the upper limit of the instrument calibration or the corresponding transmittance is lower than the lower limit of the instrument calibration, and the absorption saturation state is determined to have occurred, a dilution and retest is performed.

8. The method according to claim 2, characterized in that, Before performing formal collection, the sample is first introduced into the pre-stabilized flow chamber for buffering and flow rate adjustment for a preset time; Once the preset current stabilization conditions are met and maintained for a preset time, the sample is switched to the formal measurement channel consisting of a long optical path flow cell or a short optical path flow cell. After the sample enters the long-path flow cell or the short-path flow cell, it first undergoes a circulation flushing and replacement process to make the replacement volume reach a preset multiple of the effective volume of the flow cell and connecting pipeline before formal integration and acquisition are started.

9. The method according to claim 1, characterized in that, The conversion of the concentration of the analyte includes: The corrected characteristic response is obtained based on the characteristic peak absorbance of the analyte and the absorbance at the background wavelength. The concentration of the analyte and the repeatability verification results are obtained according to the corresponding pre-stored calibration parameters and the current dilution ratio.

10. The method according to claim 5, characterized in that, When the concentration result is invalid, a cleaning and recovery step is performed, including: The cleaning level is determined based on the residual contamination level, the contamination parameters of the previous sample, whether crystallization or wall adhesion occurs, and the blank background deviation and rinsing recovery deviation after cleaning. When the cleaning level is low, perform a single round of pure water rinsing and a blank confirmation. When the cleaning level is medium, perform pure water rinsing and low-speed replacement, and perform two blank checks. When the cleaning level is high, perform enhanced cleaning and blank verification. The wall contamination deviation is calculated by extracting the spectral difference of the wall contamination sensitive band. The wall contamination sensitive band refers to the preset characteristic band that is affected by water-soluble fertilizer deposits and whose absorbance changes exceed the preset deviation threshold. The cleaning result is determined based on the blank background deviation, the rinsing recovery deviation and the wall contamination deviation. When the cleaning result meets the release requirements, the sample injection path is opened.

Images