An intelligent titration analysis system based on artificial intelligence

Abstract

This invention relates to the field of automated titration analysis technology, specifically to an intelligent titration analysis system based on artificial intelligence. This system includes an error identification module, a temperature control module, a viscosity compensation module, a flow rate adjustment module, and a bubble correction module. By independently quantifying errors and identifying key parameter shifts in real time, this invention significantly improves the data reliability of the titration process and reduces error interference. Dynamic adjustment to cope with temperature changes, real-time adjustment of reagent viscosity and flow rate fluctuations, and identification and correction of bubble interference and density shifts all ensure the accuracy and consistency of the experiment under various environments, improving the adaptability, stability, and repeatability of titration analysis, and guaranteeing high-precision and traceable experimental results.

Description

Technical Field

[0001] This invention relates to the field of automated titration analysis technology, and in particular to an intelligent titration analysis system based on artificial intelligence. Background Technology

[0002] The field of automated titration analysis technology encompasses analytical techniques based on chemical reactions. By controlling the amount of titrant added and combining it with endpoint indication methods, the concentration or other properties of the analyte can be determined. The core components of this technology include titration methods, endpoint determination methods, reagent control, and data acquisition and processing. Titration methods mainly include acid-base titration and redox titration, each tailored to different types of analytes. Endpoint determination methods cover indicator color change, potentiometric titration, and spectroscopic titration, utilizing changes in physical or chemical signals to determine the titration endpoint. Reagent control involves flow regulation, titration speed control, and micro-injection devices to ensure accurate reagent addition. Data acquisition and processing includes real-time acquisition and storage of data such as potential, current, and absorbance during the experiment, providing a foundation for subsequent analysis. Automated titration analysis technology is widely used in chemical laboratories, environmental monitoring, food testing, and pharmaceutical analysis, achieving efficient and accurate analytical determination through computer control systems, sensors, and automated titration devices.

[0003] Among them, the AI-based intelligent titration analysis system refers to a system that combines artificial intelligence algorithms with automated titration analysis technology to achieve parameter optimization, intelligent endpoint determination, and automatic data analysis in the titration process. This system addresses issues such as reagent addition accuracy, endpoint determination error, and data processing complexity during titration. It uses machine learning models to train titration process parameters, adjusts the reagent titration rate and endpoint determination threshold based on historical experimental data, optimizes endpoint identification accuracy through neural network algorithms analyzing titration curve characteristics, and employs real-time data stream processing technology for dynamic analysis and anomaly detection of experimental data. The system regulates the reagent dropping rate through a high-precision flow control device, adaptively adjusts using potential or spectral signal feedback, and stores experimental process data and titration parameters in a database system, providing traceable analytical basis for subsequent experiments.

[0004] Existing technologies have limitations in identifying and quantifying error sources during titration. They typically rely on adjusting a single parameter or fixed threshold, making it difficult to independently optimize for different error sources. This leads to the accumulation of errors, affecting experimental accuracy. The effects of temperature changes are often ignored or compensated for with lag, making reagent flow rates susceptible to fluctuations in ambient temperature and causing endpoint drift. Changes in reagent viscosity are not adequately considered, potentially leading to unstable flow control and deviations in titration rates across different batches. Short-term flow fluctuations cannot be effectively suppressed, increasing endpoint determination errors, especially in high-precision experiments. The lack of systematic compensation for density changes and bubble interference easily introduces instability during titration, affecting the accuracy and consistency of measurement results. These shortcomings limit the adaptability of existing technologies in complex experimental environments and make it difficult to meet the demands of high-precision analysis. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of existing technologies and propose an intelligent titration analysis system based on artificial intelligence.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: an intelligent titration analysis system based on artificial intelligence includes:

[0007] The error identification module acquires flow rate, electrode potential offset, endpoint error, and temperature; calculates the offset; filters significant parameters; analyzes error correlation; filters key error sources; and obtains the error impact ranking value.

[0008] The temperature control module calls the error impact sorting value, obtains the ambient and reagent temperatures, calculates the gradient, determines the rate threshold, adjusts the pump flow rate, corrects the drive voltage, and obtains the flow rate value after temperature correction.

[0009] The viscosity compensation module calls the temperature-corrected flow rate value, obtains the reagent viscosity reference and real-time viscosity, calculates the offset and error, adjusts the pump pulse frequency, and obtains the viscosity-corrected flow rate value.

[0010] The flow adjustment module calls the viscosity-corrected flow value, obtains the flow reference and real-time value, calculates the change gradient, determines if it exceeds the limit, calculates the recovery offset value, corrects the titration flow, and obtains the flow stability adjustment value.

[0011] The bubble correction module calls the flow stability adjustment value, obtains the density reference and real-time density, calculates the offset, judges bubble interference, adjusts the driving voltage, and obtains the analysis and correction results.

[0012] As a further aspect of the present invention, the error impact ranking values ​​include the ranking of significant offset parameters, the ranking of error sources, and the correlation strength index; the temperature-corrected flow rate value specifically includes the flow rate adjustment value, the corrected driving voltage, and the adjustment effect evaluation; the viscosity-corrected flow rate value includes the adjusted pulse frequency, the correction effect evaluation, and the viscosity adjustment feedback; the flow rate stability adjustment value specifically includes the flow rate recovery efficiency, the fluctuation control effect, and the stability improvement index; and the analysis and correction results include the density correction value, the bubble impact evaluation, and the accuracy improvement after correction.

[0013] As a further aspect of the present invention, the error identification module includes:

[0014] The flow rate and potential offset calculation submodule acquires the titration reagent flow rate and electrode potential offset, calculates the flow rate offset and potential offset, calls the ambient temperature data, and calculates the offset ratio based on the error reference range using the formula:

[0015] ;

[0016] The offset ratio is calculated, and the endpoint determination error is analyzed.

[0017] in, Represents the offset ratio. This represents the flow rate of the titrant being measured. Represents the baseline flow rate. Represents the measured electrode potential. Represents the reference potential. Represents ambient temperature. Represents the reference ambient temperature.

[0018] The endpoint determination error analysis submodule calls the offset ratio to analyze the endpoint determination error, obtain the endpoint determination error offset, and calculates the endpoint determination error offset ratio based on the error benchmark range to obtain the endpoint error ratio.

[0019] The key error source screening submodule calls the endpoint error ratio to analyze the correlation of changes in differentiated error factors, screens parameters with significant deviations, identifies key error sources based on correlation analysis, and obtains the error impact ranking value.

[0020] As a further aspect of the present invention, the temperature regulation module includes

[0021] The temperature gradient calculation submodule calls the error influence sorting value to obtain the laboratory ambient temperature and reagent temperature baseline, calculates the temperature gradient, acquires temperature data at multiple time points based on temperature sampling points during the experiment, calculates the temperature change amplitude between adjacent time points, normalizes the data according to a set time interval, and calculates the temperature gradient using the formula:

[0022] ;

[0023] The temperature change gradient is calculated, and the rate of change is determined.

[0024] in, Represents the temperature gradient. This represents the temperature value at time point j. This represents the temperature value at the previous time point. Represents the total number of sampling points. Represents a time interval. Represents the laboratory ambient temperature. Represents the temperature reference of the reagent;

[0025] The flow rate adjustment submodule calls the temperature change gradient, sets the threshold range, determines whether the rate exceeds the threshold range, adjusts the flow rate of the titration reagent pump according to the set flow rate adjustment standard, calculates the flow rate offset after adjustment, obtains the flow rate adjustment offset, and corrects the titration pump drive voltage.

[0026] The temperature-corrected flow rate calculation submodule calls the flow rate adjustment offset, corrects the titration pump drive voltage based on the offset, calculates the temperature-corrected flow rate value, and obtains the temperature-corrected flow rate value.

[0027] As a further aspect of the present invention, the viscosity compensation module includes:

[0028] The viscosity offset calculation submodule calls the temperature-corrected flow rate value, obtains the titration reagent viscosity reference and real-time reagent viscosity, calculates the viscosity offset, records reagent viscosity values ​​at multiple time points based on measurement data during the experiment, and calculates the deviation from the reference viscosity using the formula:

[0029] ;

[0030] The viscosity offset is calculated, and the flow error is then calculated.

[0031] in, Represents viscosity offset. This represents the reagent viscosity measured at the p-th time point. Represents the viscosity standard of the reagent. Represents the number of measurements;

[0032] The flow error adjustment submodule calls the viscosity offset to calculate the flow error caused by viscosity change. Based on the liquid flow characteristics, it calculates the offset magnitude of viscosity change on flow through the flow influence ratio correction formula to obtain the flow error caused by viscosity change.

[0033] The pulse frequency correction submodule calls the viscosity change flow error, adjusts the pulse frequency of the titration pump, adjusts the pulse control signal according to the correction amount, dynamically compensates for the flow deviation, and obtains the flow value after viscosity correction.

[0034] As a further aspect of the present invention, the flow adjustment module includes:

[0035] The flow rate change gradient calculation submodule calls the viscosity-corrected flow rate value, obtains the standard titration flow rate reference and real-time reagent flow rate, calculates the instantaneous flow rate change gradient index, calculates the flow rate change amplitude at adjacent time points based on the flow rate measurement values ​​at different time points during the experiment, and calculates the flow rate change gradient by combining the time interval normalization, using the formula:

[0036] ;

[0037] The instantaneous change gradient index of the flow rate is calculated to determine the fluctuation.

[0038] in, Represents the gradient index of instantaneous changes in flow rate. This represents the flow rate at time point r. This represents the traffic value at the previous point in time. Represents the total number of sampling points. Represents a time interval. Represents the standard titration flow rate reference. This represents the average real-time reagent flow rate;

[0039] The traffic recovery offset calculation submodule calls the instantaneous traffic change gradient index to determine whether the fluctuation exceeds the range. Based on the traffic change threshold set in the experiment, it determines whether the instantaneous traffic fluctuation exceeds the set range and calculates the traffic recovery offset value.

[0040] The flow stabilization correction submodule calls the flow recovery offset value to correct the titration flow rate, adjust the output stability of the titration pump, adjust the pulse drive signal, dynamically compensate for flow fluctuations, and obtain the flow stabilization adjustment value.

[0041] As a further aspect of the present invention, the bubble correction module includes:

[0042] The bubble interference determination submodule obtains the reagent density benchmark and real-time reagent density based on the flow stability adjustment value. By comparing the difference between the real-time reagent density and the benchmark density, it analyzes whether there are abnormal fluctuations in the reagent density, calculates the average value of the density change amplitude, sets a density abnormality threshold, and determines whether there is bubble interference. If the density change amplitude exceeds the set threshold, it marks that there is bubble interference in the current reagent flow; otherwise, it determines that the reagent flow is normal and obtains the bubble interference determination result.

[0043] The density offset calculation submodule calls the flow rate stabilization adjustment value to obtain the reagent density baseline and real-time reagent density. It selects density data over a period of time, compares the measured real-time reagent density with the reagent density baseline, and calculates the density offset using the absolute deviation mean calculation method, employing the formula:

[0044] ;

[0045] The density offset is calculated, and error correction is performed.

[0046] in, Represents density offset. Representing the The reagent density measured in the second test, Represents the reagent density standard. Represents the number of measurements;

[0047] The flow error correction submodule determines the source of flow error based on the density offset and the bubble interference determination result. If bubble interference exists, it calculates the instantaneous flow error caused by the bubble. If there is no bubble interference, it calculates the instantaneous flow error caused by the density change, corrects the titration pump drive voltage, adjusts the titration flow rate, and obtains the analysis correction result.

[0048] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0049] In this invention, by independently quantifying errors and identifying key parameter shifts in real time, the data reliability of the titration process is significantly improved and error interference is reduced. Dynamic adjustment to cope with temperature changes, real-time adjustment of reagent viscosity and flow fluctuations, and identification and correction of bubble interference and density shifts all ensure the accuracy and consistency of the experiment under various environments, improve the adaptability, stability and repeatability of titration analysis, and guarantee high-precision and traceable experimental results. Attached Figure Description

[0050] Figure 1 This is a system flowchart of the present invention;

[0051] Figure 2 This is a flowchart of the error identification module of the present invention;

[0052] Figure 3 This is a flowchart of the temperature regulation module of the present invention;

[0053] Figure 4 This is a flowchart of the viscosity compensation module of the present invention;

[0054] Figure 5 This is a flowchart of the flow adjustment module of the present invention;

[0055] Figure 6This is a flowchart of the bubble correction module of the present invention. Detailed Implementation

[0056] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0057] 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.

[0058] Please see Figure 1 An intelligent titration analysis system based on artificial intelligence includes:

[0059] The error identification module acquires the titration reagent flow rate, electrode potential shift, endpoint determination error, and ambient temperature; calculates the flow rate shift, potential shift, and endpoint determination error shift; calculates the shift ratio based on the error reference range; calls up the parameter change trajectory of the titration process; filters parameters with significant shifts; analyzes the correlation of changes in differential error factors; filters key error sources based on the correlation; and obtains the error impact ranking value.

[0060] The temperature control module calls the error impact sorting value, obtains the laboratory ambient temperature and reagent temperature reference, calculates the temperature change gradient, sets and judges whether the rate exceeds the threshold range, adjusts the flow rate of the titration reagent pump, calculates the flow offset after adjustment, corrects the titration pump drive voltage based on the offset, and obtains the flow value after temperature correction.

[0061] The viscosity compensation module calls the flow rate value after temperature correction, obtains the viscosity reference and real-time reagent viscosity of the titration reagent, calculates the viscosity offset and flow rate influence ratio, calculates the flow rate error caused by viscosity change, adjusts the pulse frequency of the titration pump, and obtains the flow rate value after viscosity correction.

[0062] The flow adjustment module calls the viscosity-corrected flow value, obtains the standard titration flow reference and real-time reagent flow, calculates the instantaneous flow change gradient index, determines whether the fluctuation exceeds the range, calculates the flow recovery offset value, corrects the titration flow, and obtains the flow stability adjustment value.

[0063] The bubble correction module calls the flow rate stabilization adjustment value, obtains the reagent density baseline and real-time reagent density, calculates the density offset, determines whether there is bubble interference, calculates the instantaneous flow rate error, adjusts the titration pump drive voltage, and obtains the analysis correction results.

[0064] The error impact ranking values ​​include the ranking of significant offset parameters, the ranking of error sources, and the correlation strength index. The flow rate values ​​after temperature correction are specifically the flow rate adjustment value, the corrected driving voltage, and the adjustment effect evaluation. The flow rate values ​​after viscosity correction include the adjusted pulse frequency, the correction effect evaluation, and the viscosity adjustment feedback. The flow rate stability adjustment values ​​are specifically the flow rate recovery efficiency, the fluctuation control effect, and the stability improvement index. The analysis of correction results includes the density correction value, the bubble impact evaluation, and the accuracy improvement after correction.

[0065] Please see Figure 2 The error identification module includes:

[0066] The flow rate and potential offset calculation submodule acquires the titration reagent flow rate and electrode potential offset, calculates the flow rate offset and potential offset, calls the ambient temperature data, and calculates the offset ratio based on the error reference range using the formula:

[0067] ;

[0068] The offset ratio is calculated, and the endpoint determination error is analyzed.

[0069] in, Represents the offset ratio. This represents the flow rate of the titrant being measured. Represents the baseline flow rate. Represents the measured electrode potential. Represents the reference potential. Represents ambient temperature. Represents the reference ambient temperature.

[0070] First, a high-precision flow meter is used to measure the actual flow rate of the titrant during the titration process. For example, if the target flow rate is set to 10.0 mL / min and the actual measured value is 10.5 mL / min, the flow rate offset is 0.5 mL / min. Simultaneously, a precision potentiometer is used to measure the actual electrode potential. Assuming the reference potential is 200.0 mV and the measured value is 205.0 mV, the potential offset is 5.0 mV. Next, ambient temperature data is collected. Assuming the laboratory temperature is 25.0°C and the reference temperature is 20.0°C, the temperature offset is 5.0°C. Based on the error reference range, the offset ratios for each item are calculated using the formula:

[0071] ;

[0072] in, This is the offset ratio. For the measurement of titration reagent flow rate, As the baseline flow rate, The measured electrode potential, As the reference potential, For ambient temperature, Using the baseline ambient temperature as an example, substituting the above values, we can calculate:

[0073] ;

[0074] Therefore, the offset ratio Approximately 2.311, representing the combined deviation of flow rate, potential, and temperature.

[0075] The endpoint determination error analysis submodule calls the offset ratio to analyze the endpoint determination error, obtain the endpoint determination error offset, and calculate the endpoint determination error offset ratio based on the error benchmark range to obtain the endpoint error ratio.

[0076] To analyze the endpoint determination error, firstly, the theoretical endpoint potential during the titration process is determined, assumed to be 400.0 mV. Then, the actual endpoint potential is measured, assumed to be 410.0 mV, and the endpoint potential error is calculated to be 10.0 mV. Next, the influence of the offset ratio on endpoint determination is considered, assuming a linear relationship between the offset ratio and the endpoint error, as shown in the following formula:

[0077] ;

[0078] in, For the endpoint potential error, Let this be the proportionality coefficient, assuming ,but:

[0079] ;

[0080] The calculated endpoint potential error is approximately 9.244 mV, which is close to the actual measured 10.0 mV, indicating that the offset ratio has a significant impact on the endpoint determination error.

[0081] The key error source screening submodule calls the endpoint error ratio, analyzes the correlation of changes in differential error factors, filters parameters with significant deviations, identifies key error sources based on correlation analysis, and obtains the error impact ranking value.

[0082] First, comparing the magnitudes of the various offsets, the flow rate offset was 0.5 mL / min, accounting for 5% of the baseline flow rate; the potential offset was 5.0 mV, accounting for 2.5% of the baseline potential; and the temperature offset was 5.0°C, with a corresponding offset ratio contribution of approximately 2.236, significantly higher than the offset ratio contributions of flow rate and potential. Therefore, changes in ambient temperature are the main source of error, and it is recommended to strictly control the ambient temperature during the experiment to reduce its impact on the titration results.

[0083] Please see Figure 3 The temperature control module includes

[0084] The temperature gradient calculation submodule calls the error impact sorting value to obtain the laboratory ambient temperature and reagent temperature baseline, calculates the temperature gradient, and obtains temperature data at multiple time points based on temperature sampling points during the experiment. It calculates the temperature change amplitude between adjacent time points, normalizes the data according to a set time interval, and calculates the temperature gradient using the formula:

[0085] ;

[0086] The temperature change gradient is calculated, and the rate of change is determined.

[0087] in, Represents the temperature gradient. This represents the temperature value at time point j. This represents the temperature value at the previous time point. Represents the total number of sampling points. Represents a time interval. Represents the laboratory ambient temperature. Represents the temperature reference of the reagent;

[0088] First, the laboratory ambient temperature is monitored at different time points using a high-precision temperature sensor. Assuming temperature data is recorded every 10 seconds, a temperature sequence is obtained:

[0089] ;

[0090] Then, determine the reagent temperature baseline, assuming the experiment requires the reagent temperature to be maintained at a certain level. The specific steps for calculating the temperature gradient are as follows:

[0091] Calculate the temperature change at adjacent time points:

[0092] ;

[0093] ;

[0094] ;

[0095] ;

[0096] Calculate the temperature gradient The values ​​are calculated using normalization:

[0097] ;

[0098] Substitute the known data:

[0099] ;

[0100] ;

[0101] ;

[0102] To determine whether the temperature change gradient exceeds the allowable range, the maximum allowable gradient is set as follows: Calculated value The flow rate exceeds the threshold range, therefore adjustment is necessary. The final temperature gradient is then obtained for subsequent rate determination.

[0103] The flow rate adjustment submodule calls the temperature change gradient, sets the threshold range, determines whether the rate exceeds the threshold range, adjusts the flow rate of the titration reagent pump according to the set flow rate adjustment standard, calculates the flow rate offset after adjustment, obtains the flow rate adjustment offset, and corrects the titration pump drive voltage.

[0104] Call the temperature change gradient, set a threshold range, and determine if the rate exceeds the threshold range. Based on experimental requirements, set the threshold range as follows: Check if the rate of temperature change exceeds this range. If it does, adjust the flow rate of the titration reagent pump, for example, the currently calculated flow rate. for If the flow rate exceeds the set threshold, it needs to be adjusted. Calculate the adjusted flow rate offset and reduce the flow rate. The correction is performed to obtain the flow adjustment offset, which is then used to correct the titration pump drive voltage.

[0105] The temperature-corrected flow calculation submodule calls the flow adjustment offset, corrects the titration pump drive voltage based on the offset, calculates the temperature-corrected flow value, and obtains the temperature-corrected flow value.

[0106] First, obtain the adjusted flow rate offset to determine whether the current flow rate meets the stability requirements of reagent titration. If the offset exceeds the set range, for example, if the experimental setting allows a flow rate offset range of ±0.3 mL / min, but the currently measured offset is 0.5 mL / min, the drive voltage needs to be adjusted. Increase or decrease the pump's power supply voltage according to the offset direction. For example, if the flow rate offset is positive, decrease the voltage; if it is negative, increase the voltage. At the same time, monitor the adjusted flow rate and continue adjusting until the flow rate offset returns to the allowable range. For example, if the adjusted flow rate offset drops to 0.2 mL / min, it meets the set standard. Finally, obtain the temperature-corrected flow rate value.

[0107] Please see Figure 4 The viscosity compensation module includes:

[0108] The viscosity offset calculation submodule calls the temperature-corrected flow rate value, obtains the titration reagent viscosity reference and real-time reagent viscosity, calculates the viscosity offset, records reagent viscosity values ​​at multiple time points based on measurement data during the experiment, and calculates the deviation from the reference viscosity using the formula:

[0109] ;

[0110] The viscosity offset is calculated, and the flow error is then calculated.

[0111] in, Represents viscosity offset. This represents the reagent viscosity measured at the p-th time point. Represents the viscosity standard of the reagent. Represents the number of measurements;

[0112] First, monitor the actual viscosity value of the titration reagent and record the dynamic viscosity of the reagent at different time points using a viscometer. For example, the reference viscosity of a certain reagent is set to... In the experiment, the viscosity values ​​were measured at multiple time points, namely mPa·s. mPa·s, mPa·s, mPa·s, mPa·s and Then, calculate the viscosity shift using the formula: mPa·s.

[0113] ;

[0114] Substitute the measurement data into the calculation:

[0115] ;

[0116] ;

[0117] The final calculated viscosity shift is mPa·s.

[0118] The flow error adjustment submodule calls the viscosity offset to calculate the flow error caused by viscosity change. Based on the liquid flow characteristics, it calculates the offset of viscosity change on flow through the flow influence ratio correction formula to obtain the flow error caused by viscosity change.

[0119] First, the flow state of the titrant is analyzed. The flow rate of the titrant is affected by viscosity. Increased viscosity leads to increased flow resistance, which in turn affects the stability of the flow rate. If the reference flow rate of the reagent is set to 5.0 mL / min, and the currently measured viscosity deviation is 0.104 mPa·s, it is necessary to evaluate the impact of this viscosity change on the flow rate and set an experimental measurement threshold range. For example, if a viscosity change of 0.1 mPa·s causes a decrease in flow rate of 0.3 mL / min, the current deviation may lead to a flow rate deviation of approximately 0.3-0.4 mL / min. The experimentally measured current flow rate is 4.65 mL / min, which is significantly different from the reference flow rate of 5.0 mL / min. Therefore, the flow rate error caused by viscosity needs to be corrected. The error value is input into the control system and compared with the set flow rate threshold to determine whether it exceeds the allowable error range. For example, if the error threshold is set to ±0.2 mL / min, the current error exceeds this range. Therefore, the pump pulse frequency needs to be adjusted to compensate for the flow loss. Finally, the flow rate error due to viscosity change is obtained, and subsequent pulse frequency adjustments are made.

[0120] The pulse frequency correction submodule calls the viscosity change flow error, adjusts the pulse frequency of the titration pump, adjusts the pulse control signal according to the correction amount, dynamically compensates for the flow deviation, and obtains the flow value after viscosity correction.

[0121] First, obtain the current pump pulse frequency setting value. For example, the reference pulse frequency is set to 100Hz. The flow error is usually adjusted by changing the pump's drive pulse frequency. If a decrease in flow is detected, the pulse frequency needs to be increased to compensate for the flow loss. Assuming the flow error is 0.35mL / min, the pulse frequency can be increased by 5-7Hz based on experience. After the adjustment, the pump's pulse frequency is set to 105Hz. Measure the titration reagent flow rate again. If the flow rate recovers to 5.0±0.1mL / min, the adjustment is complete. If the flow rate still deviates from the set range, continue to fine-tune the pulse frequency until the flow rate reaches a stable state. Finally, obtain the viscosity-corrected flow rate value.

[0122] Please see Figure 5 The flow adjustment module includes:

[0123] The flow rate change gradient calculation submodule calls the viscosity-corrected flow rate value, obtains the standard titration flow rate reference and real-time reagent flow rate, calculates the instantaneous flow rate change gradient index, calculates the flow rate change amplitude at adjacent time points based on the flow rate measurements at different time points during the experiment, and calculates the flow rate change gradient by combining time interval normalization, using the formula:

[0124] ;

[0125] The instantaneous change gradient index of the flow rate is calculated to determine the fluctuation.

[0126] in, Represents the gradient index of instantaneous changes in flow rate. This represents the flow rate at time point r. This represents the traffic value at the previous point in time. Represents the total number of sampling points. Represents a time interval. Represents the standard titration flow rate reference. This represents the average real-time reagent flow rate;

[0127] First, set the total number of samples. Interval between each measurement The traffic data was standardized, and then the instantaneous traffic change gradient was calculated using the following formula:

[0128] ;

[0129] During the calculation process, the total number of samples is set. Measurement time interval Collect flow rate data, assuming the measured reagent flow rates are as follows: , ... (Unit: mL / s), calculate the flow rate change at adjacent time points:

[0130] ;

[0131] Calculate the normalized gradient:

[0132] ;

[0133] Calculate the mean deviation term:

[0134] ;

[0135] ;

[0136] Finally, the gradient of instantaneous flow rate change is calculated:

[0137] ;

[0138] Set judgment threshold ,because If the flow fluctuation is abnormal, adjustments are needed to ultimately obtain the instantaneous flow change gradient index.

[0139] The traffic recovery offset calculation submodule calls the instantaneous traffic change gradient index to determine whether the fluctuation exceeds the range. Based on the traffic change threshold set in the experiment, it determines whether the instantaneous traffic fluctuation exceeds the set range and calculates the traffic recovery offset value.

[0140] First, set the allowable flow rate fluctuation range, for example, ±0.2 mL / min. The measured data shows a maximum fluctuation of 0.8 mL / min, exceeding the set range. Therefore, calculate the flow rate recovery offset value, determine the corrected target value, and calculate the offset. For example, if the baseline flow rate is 5.0 mL / min, and the measured average is 5.06 mL / min, then the offset is calculated as follows:

[0141] ;

[0142] Since the offset is small, it can be left unadjusted for the time being. However, if the offset exceeds 0.2 mL / min, further correction is required to finally obtain the flow recovery offset value.

[0143] The flow stabilization correction submodule calls the flow recovery offset value, corrects the titration flow rate, adjusts the output stability of the titration pump, adjusts the pulse drive signal, dynamically compensates for flow fluctuations, and obtains the flow stabilization adjustment value.

[0144] First, adjust the pulse drive signal according to the offset. For example, if the current pump pulse frequency is 100Hz, and the flow fluctuation is detected to exceed 0.2mL / min, adjust the pulse frequency by ±5Hz. If the flow rate stabilizes at 5.0mL / min ±0.1mL / min after adjustment, the adjustment is complete. If there is still a deviation, continue to fine-tune the pulse frequency to finally obtain a stable flow rate adjustment value.

[0145] Please see Figure 6 The bubble correction module includes:

[0146] The bubble interference determination submodule obtains the reagent density baseline and real-time reagent density based on the flow stability adjustment value. By comparing the difference between the real-time reagent density and the baseline density, it analyzes whether there are abnormal fluctuations in the reagent density, calculates the average value of the density change amplitude, sets a density abnormality threshold, and determines whether there is bubble interference. If the density change amplitude exceeds the set threshold, it marks the current reagent flow as having bubble interference; otherwise, it determines that the reagent flow is normal and obtains the bubble interference determination result.

[0147] First, the flow rate stabilization adjustment value is called to extract recent reagent flow rate variation data and analyze it in conjunction with the reagent density benchmark. A density measurement time window is set, and reagent density data is periodically collected. The reagent density at different time points is normalized, and the deviation value from the reagent density benchmark is calculated. The density offset is modeled as a time series to detect its trend characteristics and determine whether there is a sudden change in density. If the density change exceeds the set normal offset threshold, bubble interference may exist. Based on this, a bubble influence model is established by combining flow rate changes. By analyzing the relationship between the density offset rate and the instantaneous flow rate change rate in a short period of time, bubble interference signals are identified. If the density changes suddenly in a short period of time and the flow rate fluctuation rate reaches the set threshold, it is judged as bubble interference. The interference time point is recorded, and the instantaneous flow error is calculated. The error correction requirement is analyzed. If the error is within the allowable range, no correction is performed. If the error exceeds the set error tolerance, the titration pump drive voltage needs to be adjusted. The adjustment range is determined by the instantaneous flow error, density offset, and historical error correction data. Finally, the analysis and correction results are obtained.

[0148] The density offset calculation submodule calls the flow rate stabilization adjustment value to obtain the reagent density baseline and real-time reagent density. It selects density data over a period of time, compares the measured real-time reagent density with the reagent density baseline, and calculates the density offset using the absolute deviation mean calculation method, employing the formula:

[0149] ;

[0150] The density offset is calculated, and error correction is performed.

[0151] in, Represents density offset. Representing the The reagent density measured in the second test, Represents the reagent density standard. Represents the number of measurements;

[0152] The measurement period needs to be determined during the density offset calculation process. During the measurement period The reagent density was collected and the measurement data were recorded. The reagent density benchmark was determined using the mean value calculation formula. Calculate the absolute deviation between each measurement and the reference value, and take the average value as the density offset:

[0153] ;

[0154] in, Represents density offset. Representing the The reagent density measured in the second test, Represents the reagent density standard. This represents the number of measurements.

[0155] Calculation steps:

[0156] Set measurement cycle Choose an appropriate measurement cycle For example, 10 samples are taken within 5 seconds, with each measurement taken at a 0.5-second interval;

[0157] Collect reagent density data: within a set period. Secondary density measurement, assuming the measured density values ​​are respectively (Unit: g / cm3);

[0158] Calculate the reagent density benchmark: Take the mean of all measurements as the density benchmark.

[0159] ;

[0160] Calculate the density offset using the formula described above:

[0161] ;

[0162] Determine if the density offset exceeds the threshold: Set the offset threshold. ,at this time If the density shift does not exceed the threshold, it is considered to be within the normal range.

[0163] The flow error correction submodule determines the source of flow error based on the density offset and bubble interference judgment results. If bubble interference exists, it calculates the instantaneous flow error caused by bubbles. If there is no bubble interference, it calculates the instantaneous flow error caused by density change, corrects the titration pump drive voltage, adjusts the titration flow rate, and obtains the analysis correction results.

[0164] First, set an error correction threshold. If the instantaneous flow error is lower than the set threshold, no adjustment is made. If it exceeds the threshold, the correction range is determined according to the error correction model. By adjusting the increase or decrease of the titration pump drive voltage, the titration flow rate is gradually brought back to a stable state. During the adjustment process, an error monitoring window is set to periodically detect the error correction. If the error still exceeds the allowable range after adjustment, the drive voltage is adjusted until the error converges to the set range.

[0165] 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. An intelligent titration analysis system based on artificial intelligence, characterized in that, The system includes: The error identification module acquires flow rate, electrode potential offset, endpoint error, and temperature; calculates the offset; filters significant parameters; analyzes error correlation; filters key error sources; and obtains the error impact ranking value. The temperature control module calls the error impact sorting value, obtains the ambient and reagent temperatures, calculates the gradient, determines the rate threshold, adjusts the pump flow rate, corrects the drive voltage, and obtains the flow rate value after temperature correction. The viscosity compensation module calls the temperature-corrected flow rate value, obtains the reagent viscosity reference and real-time viscosity, calculates the offset and error, adjusts the pump pulse frequency, and obtains the viscosity-corrected flow rate value. The flow adjustment module calls the viscosity-corrected flow value, obtains the flow reference and real-time value, calculates the change gradient, determines if it exceeds the limit, calculates the recovery offset value, corrects the titration flow, and obtains the flow stability adjustment value. The bubble correction module calls the flow stability adjustment value, obtains the density reference and real-time density, calculates the offset, judges bubble interference, adjusts the driving voltage, and obtains the analysis and correction results.

2. The intelligent titration analysis system based on artificial intelligence according to claim 1, characterized in that, The error impact ranking values ​​include the ranking of significant offset parameters, the ranking of error sources, and the correlation strength index. The temperature-corrected flow rate value specifically includes the flow rate adjustment value, the corrected driving voltage, and the adjustment effect evaluation. The viscosity-corrected flow rate value includes the adjusted pulse frequency, the correction effect evaluation, and the viscosity adjustment feedback. The flow rate stability adjustment value specifically includes the flow rate recovery efficiency, the fluctuation control effect, and the stability improvement index. The analysis and correction results include the density correction value, the bubble impact evaluation, and the accuracy improvement after correction.

3. The intelligent titration analysis system based on artificial intelligence according to claim 1, characterized in that, The error identification module includes: The flow rate and potential offset calculation submodule acquires the titration reagent flow rate and electrode potential offset, calculates the flow rate offset and potential offset, calls the ambient temperature data, and calculates the offset ratio based on the error reference range using the formula: ； The offset ratio is calculated, and the endpoint determination error is analyzed. in, Represents the offset ratio. This represents the flow rate of the titrant being measured. Represents the baseline flow rate. Represents the measured electrode potential. Represents the reference potential. Represents ambient temperature. Represents the reference ambient temperature; The endpoint determination error analysis submodule calls the offset ratio to analyze the endpoint determination error, obtain the endpoint determination error offset, and calculates the endpoint determination error offset ratio based on the error benchmark range to obtain the endpoint error ratio. The key error source screening submodule calls the endpoint error ratio to analyze the correlation of changes in differentiated error factors, screens parameters with significant deviations, identifies key error sources based on correlation analysis, and obtains the error impact ranking value.

4. The intelligent titration analysis system based on artificial intelligence according to claim 1, characterized in that, The temperature regulation module includes The temperature gradient calculation submodule calls the error influence sorting value to obtain the laboratory ambient temperature and reagent temperature baseline, calculates the temperature gradient, acquires temperature data at multiple time points based on temperature sampling points during the experiment, calculates the temperature change amplitude between adjacent time points, normalizes the data according to a set time interval, and calculates the temperature gradient using the formula: ； The temperature change gradient is calculated, and the rate of change is determined. in, Represents the temperature gradient. This represents the temperature value at time point j. This represents the temperature value at the previous time point. Represents the total number of sampling points. Represents a time interval. Represents the laboratory ambient temperature. Represents the temperature reference of the reagent; The flow rate adjustment submodule calls the temperature change gradient, sets the threshold range, determines whether the rate exceeds the threshold range, adjusts the flow rate of the titration reagent pump according to the set flow rate adjustment standard, calculates the flow rate offset after adjustment, obtains the flow rate adjustment offset, and corrects the titration pump drive voltage. The temperature-corrected flow rate calculation submodule calls the flow rate adjustment offset, corrects the titration pump drive voltage based on the offset, calculates the temperature-corrected flow rate value, and obtains the temperature-corrected flow rate value.

5. The intelligent titration analysis system based on artificial intelligence according to claim 1, characterized in that, The viscosity compensation module includes: The viscosity offset calculation submodule calls the temperature-corrected flow rate value, obtains the titration reagent viscosity reference and real-time reagent viscosity, calculates the viscosity offset, records reagent viscosity values ​​at multiple time points based on measurement data during the experiment, and calculates the deviation from the reference viscosity using the formula: ； The viscosity offset is calculated, and the flow error is then calculated. in, Represents viscosity offset. This represents the reagent viscosity measured at the p-th time point. Represents the viscosity standard of the reagent. Represents the number of measurements; The flow error adjustment submodule calls the viscosity offset to calculate the flow error caused by viscosity change. Based on the liquid flow characteristics, it calculates the offset magnitude of viscosity change on flow through the flow influence ratio correction formula to obtain the flow error caused by viscosity change. The pulse frequency correction submodule calls the viscosity change flow error, adjusts the pulse frequency of the titration pump, adjusts the pulse control signal according to the correction amount, dynamically compensates for the flow deviation, and obtains the flow value after viscosity correction.

6. The intelligent titration analysis system based on artificial intelligence according to claim 1, characterized in that, The flow adjustment module includes: The flow rate change gradient calculation submodule calls the viscosity-corrected flow rate value, obtains the standard titration flow rate reference and real-time reagent flow rate, calculates the instantaneous flow rate change gradient index, calculates the flow rate change amplitude at adjacent time points based on the flow rate measurement values ​​at different time points during the experiment, and calculates the flow rate change gradient by combining the time interval normalization, using the formula: ; The instantaneous change gradient index of the flow rate is calculated to determine the fluctuation. in, Represents the gradient index of instantaneous changes in flow rate. This represents the flow rate at time point r. This represents the traffic value at the previous point in time. Represents the total number of sampling points. Represents a time interval. Represents the standard titration flow rate reference. This represents the average real-time reagent flow rate; The traffic recovery offset calculation submodule calls the instantaneous traffic change gradient index to determine whether the fluctuation exceeds the range. Based on the traffic change threshold set in the experiment, it determines whether the instantaneous traffic fluctuation exceeds the set range and calculates the traffic recovery offset value. The flow stabilization correction submodule calls the flow recovery offset value to correct the titration flow rate, adjust the output stability of the titration pump, adjust the pulse drive signal, dynamically compensate for flow fluctuations, and obtain the flow stabilization adjustment value.

7. The intelligent titration analysis system based on artificial intelligence according to claim 1, characterized in that, The bubble correction module includes: The bubble interference determination submodule obtains the reagent density benchmark and real-time reagent density based on the flow stability adjustment value. By comparing the difference between the real-time reagent density and the benchmark density, it analyzes whether there are abnormal fluctuations in the reagent density, calculates the average value of the density change amplitude, sets a density abnormality threshold, and determines whether there is bubble interference. If the density change amplitude exceeds the set threshold, it marks that there is bubble interference in the current reagent flow; otherwise, it determines that the reagent flow is normal and obtains the bubble interference determination result. The density offset calculation submodule calls the flow rate stabilization adjustment value to obtain the reagent density baseline and real-time reagent density. It selects density data over a period of time, compares the measured real-time reagent density with the reagent density baseline, and calculates the density offset using the absolute deviation mean calculation method, employing the formula: ； The density offset is calculated, and error correction is performed. in, Represents density offset. Representing the The reagent density measured in the second test, Represents the reagent density standard. Represents the number of measurements; The flow error correction submodule determines the source of flow error based on the density offset and the bubble interference determination result. If bubble interference exists, it calculates the instantaneous flow error caused by the bubble. If there is no bubble interference, it calculates the instantaneous flow error caused by the density change, corrects the titration pump drive voltage, adjusts the titration flow rate, and obtains the analysis correction result.

Images