Deep learning based prediction of potentiometric titration end point and outlier correction method

Abstract

The application relates to the technical field of potential titration prediction and correction, in particular to a potential titration end point prediction and abnormal value correction method based on deep learning, which comprises the following steps: acquiring potential change data in a potential titration process, basic data of a titrant and parameters of a sample to be measured; determining whether it is a suspected abnormal potential according to whether the potential change slope exceeds a preset value and / or whether the potential change violates monotonicity; predicting the abnormal potential type based on the mutation amplitude of the suspected abnormal potential and the preset amplitude comparison; predicting whether it is a weak abnormal potential by whether the local curvature change rate of the weak abnormal potential adjacent potential exceeds a preset value and the time sequence correlation; and determining the correction method according to the comparison result of the abnormal potential duration, the abnormal value deviation degree and the preset value. The application improves the accuracy of the weak abnormal condition judgment of the potential titration technology, and further reduces the error of the analysis result.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of potential titration prediction and correction, and particularly relates to a potential titration end point prediction and abnormal value correction method based on deep learning. BACKGROUND

[0002] In the field of potential titration analysis, the traditional method relies on manual observation of potential jump or judgment of titration end point based on simple first-order derivative calculation, which has the problems of poor end point judgment accuracy and easy interference by human factors. At the same time, abnormal potential data may be generated in the experimental process due to electrode fluctuation, environmental noise, instrument failure, etc. The existing technology lacks effective multi-dimensional abnormality detection and adaptive correction mechanism, and it is difficult to distinguish between strong and weak abnormalities, and the processing capacity for complex abnormal conditions is insufficient, resulting in low repeatability and large error of analysis results. With the increasing requirements of chemical, environmental, pharmaceutical and other industries on analysis accuracy and automation level, there is an urgent need for a potential titration method that can realize real-time abnormality detection, accurate end point prediction and intelligent data correction.

[0003] For example, Chinese patent application publication No. CN114414648A discloses an automatic potential titration method and system based on machine learning, which relates to the technical field of automatic potential titration. The method comprises the following steps: obtaining the initial potential of the solution to be measured; identifying the label information of the burette containing the titrant, the label information including the titration concentration of the titrant; predicting the titration end point based on the titration concentration and the initial potential of the solution and through a prediction model; performing titration analysis on the solution to be measured through the burette, and collecting real-time titration data of the solution to be measured during titration analysis; determining whether the titration analysis meets the termination condition based on the real-time titration data and the titration end point; if the titration analysis meets the termination condition, terminating the titration analysis, and calculating the ion concentration of the ion to be measured in the solution to be measured according to the real-time titration data at the time of termination of the titration analysis. The application has the effect of improving the ion concentration detection accuracy by using a prediction model based on machine learning to predict the titration end point.

[0004] However, the existing technology has the problem that the potential titration technology cannot accurately judge weak abnormal conditions, resulting in large errors in analysis results. SUMMARY

[0005] Therefore, the present application provides a potential titration end point prediction and abnormal value correction method based on deep learning to overcome the problem in the prior art that the potential titration technology cannot accurately judge weak abnormal conditions, resulting in large errors in analysis results.

[0006] To achieve the above-mentioned purpose, the present application provides a potential titration end point prediction and abnormal value correction method based on deep learning, comprising:

[0007] Step S1, obtaining potential change data in the potential titration process, basic data of the titrant, and parameters of the sample to be measured;

[0008] In step S2, it is determined whether the potential is a suspected abnormal potential based on whether the potential change slope exceeds a preset change slope and / or whether the potential change violates the monotonicity of the potential change;

[0009] In step S3, the abnormal type of the suspected abnormal potential is predicted based on a comparison result of the mutation amplitude of the suspected abnormal potential with the first preset mutation amplitude and the second preset mutation amplitude;

[0010] In step S4, it is determined whether the current potential is a weak abnormal potential or a normal potential based on whether the local curvature change rate of the adjacent potential of the weak abnormal potential titration is greater than a preset curvature change rate and a time sequence correlation;

[0011] In step S5, the abnormal potential titration is corrected by different correction methods based on a comparison result of the duration of the abnormal potential with a preset duration and a comparison result of the deviation degree of the abnormal potential titration abnormal value with a preset deviation degree.

[0012] Further, in the step S2, it is determined whether the potential is a suspected abnormal potential based on whether the potential change slope exceeds a preset change slope and whether the potential change violates the monotonicity of the potential change, comprising:

[0013] If the potential change slope exceeds the preset change slope or the potential change violates the monotonicity of the potential change, it is determined that the potential is a suspected abnormal potential;

[0014] If the potential change slope does not exceed the preset change slope and the potential change conforms to the monotonicity of the potential change, it is determined that the potential is a normal potential.

[0015] Further, the potential change slope is determined according to the ratio of the potential difference of the adjacent two sampling points to the difference of the titrant addition amount, the preset change slope is determined according to the average potential change slope of a plurality of same type potential titrations, and the monotonicity of the potential change is determined according to the potential value change trend after the titrant is added.

[0016] Further, in the step S3, the abnormal type of the suspected abnormal potential is predicted based on the suspected abnormal potential and a comparison result of the mutation amplitude of the suspected abnormal potential with the first preset mutation amplitude and the second preset mutation amplitude, comprising:

[0017] If the mutation amplitude of the suspected abnormal potential is greater than the first preset mutation amplitude, it is predicted that the abnormal type of the suspected abnormal potential is a strong abnormal potential;

[0018] If the mutation amplitude of the suspected abnormal potential is less than or equal to the first preset mutation amplitude and greater than the second preset mutation amplitude, it is predicted that the abnormal type of the suspected abnormal potential is a weak abnormal potential.

[0019] If the mutation amplitude of the suspected abnormal potential is less than or equal to the second preset mutation amplitude, it is predicted that the potential is not abnormal.

[0020] Further, the mutation amplitude of the suspected abnormal potential is determined according to the absolute value of the potential difference between the suspected abnormal potential and the adjacent normal point, the first preset mutation amplitude is determined according to the maximum amplitude of the normal fluctuation in the historical experimental data, and the second preset mutation amplitude is half of the first preset amplitude.

[0021] Further, in the step S4, whether the local curvature change rate of the adjacent potential of the weak abnormal potential titration is greater than the preset curvature change rate and the time sequence correlation is predicted to determine whether the current potential is a weak abnormal potential or a normal potential includes:

[0022] If the local curvature change rate of the adjacent potential of the weak abnormal potential titration is greater than the preset curvature change rate and the time sequence correlation of the weak abnormal potential is less than the preset time sequence correlation, it is predicted that the current potential is a weak abnormal potential.

[0023] If the local curvature change rate of the adjacent potential of the weak abnormal potential titration is less than or equal to the preset curvature change rate and the time sequence correlation of the weak abnormal potential is greater than or equal to the preset time sequence correlation, it is predicted that the current potential is a normal potential.

[0024] Further, the local curvature change rate of the adjacent potential of the weak abnormal potential titration is determined according to the first and second derivatives of the potential with respect to time, and the preset curvature change rate is determined according to the average local curvature change rate of the adjacent potential of the same type of weak abnormal potential titration.

[0025] Further, the time sequence correlation is determined according to the dynamic time warping distance, and the preset time sequence correlation is determined according to the time sequence correlation distribution of all sequences in the historical normal data.

[0026] Further, in the step S5, based on whether the duration of the abnormal potential is greater than the preset duration and the comparison result of the deviation degree of the abnormal value of the abnormal potential titration and the preset deviation degree, the abnormal potential titration is corrected by different correction methods includes:

[0027] If the duration of the abnormal potential is less than the preset duration and the deviation degree of the abnormal value is less than the preset deviation degree, it is determined that the abnormal potential titration does not need to be corrected.

[0028] If the duration of the abnormal potential is less than the preset duration and the deviation degree of the abnormal value of the abnormal potential titration is greater than or equal to the preset deviation degree, it is determined that the abnormal potential titration is corrected by the K-nearest neighbor algorithm.

[0029] If the duration of the abnormal potential is greater than or equal to the preset duration and the deviation degree of the abnormal potential titration abnormal value is less than the preset deviation degree, it is determined that the abnormal potential titration is corrected by the window sliding method.

[0030] If the duration of the abnormal potential is greater than or equal to the preset duration and the deviation degree of the abnormal potential titration abnormal value is greater than or equal to the preset deviation degree, it is determined that the potential titration is re-prepared.

[0031] Further, the preset duration is determined according to the historical average duration of the abnormal potential under the standard state, the deviation degree of the abnormal value is determined according to the absolute value of the potential difference between the abnormal value and the ideal titration curve at the corresponding point, and the preset deviation degree is determined according to the average deviation degree under the standard state of several same type potential titrations.

[0032] Compared with the prior art, the beneficial effects of the present application are that, by determining whether the potential change slope exceeds the preset change slope and whether the potential change violates the monotonicity of the potential change, it is determined whether the potential is a suspected abnormal potential, according to the potential change slope exceeding the preset change slope or the potential change violating the monotonicity of the potential change, it is explained that the potential mutation or trend abnormal phenomenon occurs, the suspected abnormal potential is accurately determined, according to the potential change slope not exceeding the preset change slope and the potential change conforming to the monotonicity of the potential change, it is explained that the potential change rate is normal and the trend is consistent, the normal potential is accurately determined, the accuracy of the weak abnormal condition judgment of the potential titration technology is improved through the above content, and the error of the analysis result is reduced.

[0033] Further, by comparing the suspected abnormal potential and the mutation amplitude of the suspected abnormal potential with the first preset mutation amplitude and the second preset mutation amplitude, the abnormal type of the suspected abnormal potential is predicted, according to the mutation amplitude of the suspected abnormal potential being greater than the first preset mutation amplitude, it is explained that the potential value jumps significantly and exceeds the upper limit of the normal fluctuation range, the abnormal type of the suspected abnormal potential is accurately predicted as a strong abnormal potential, according to the mutation amplitude of the suspected abnormal potential being less than or equal to the first preset mutation amplitude and greater than the second preset mutation amplitude, it is explained that the potential value exists perceptible abnormal fluctuation but does not reach the significant jump degree, the abnormal type of the suspected abnormal potential is accurately predicted as a weak abnormal potential, and according to the mutation amplitude of the suspected abnormal potential being less than or equal to the second preset mutation amplitude, it is explained that the potential fluctuation is within the normal noise range or only exists the inherent error of the instrument, the potential is accurately predicted as not being abnormal, the accuracy of the weak abnormal condition judgment of the potential titration technology is improved through the above content, and the error of the analysis result is reduced.

[0034] Further, the present application predicts the current potential as a weak abnormal potential or a normal potential by whether the local curvature change rate of the adjacent potential of the weak abnormal potential titration is greater than the preset curvature change rate and the timing correlation, according to the local curvature change rate of the adjacent potential of the weak abnormal potential titration being greater than the preset curvature change rate and the timing correlation of the weak abnormal potential being less than the preset timing correlation, it is indicated that the local bending degree of the potential curve exceeds the normal range, and the fluctuation mode is different from the historical normal sequence, accurately predicting the current potential as a weak abnormal potential, according to the local curvature change rate of the adjacent potential of the weak abnormal potential titration being less than or equal to the preset curvature change rate and the timing correlation of the weak abnormal potential being greater than or equal to the preset timing correlation, it is indicated that the potential curve is smooth and the fluctuation mode is highly similar to the historical normal sequence, accurately predicting the current potential as a normal potential, through the above content, the accuracy of the weak abnormal situation judgment of the potential titration technology is improved, and the error of the analysis result is reduced.

[0035] Further, the present application determines to correct the abnormal potential titration by different correction methods according to whether the duration of the abnormal potential is greater than the preset duration and the comparison result of the deviation degree of the abnormal value of the abnormal potential titration and the preset deviation degree, according to the duration of the abnormal potential being less than the preset duration and the deviation degree of the abnormal value being less than the preset deviation degree, it is indicated that the normal noise of slight fluctuation and self-recovery phenomenon occurs, accurately determining that the abnormal potential titration does not need to be corrected, according to the duration of the abnormal potential being less than the preset duration and the deviation degree of the abnormal value of the abnormal potential titration being greater than or equal to the preset deviation degree, it is indicated that the isolated abnormal phenomenon of sudden strong interference but not lasting occurs, accurately determining that the abnormal potential titration is corrected by K nearest neighbor algorithm, according to the duration of the abnormal potential being greater than or equal to the preset duration and the deviation degree of the abnormal value of the abnormal potential titration being less than the preset deviation degree, it is indicated that the systematic drift phenomenon of long-term existence but small amplitude occurs, accurately determining that the abnormal potential titration is corrected by the window sliding method, according to the duration of the abnormal potential being greater than or equal to the preset duration and the deviation degree of the abnormal value of the abnormal potential titration being greater than or equal to the preset deviation degree, it is indicated that the fatal abnormal phenomenon of serious and persistent occurs, accurately determining that the potential titration is prepared again, through the above content, the accuracy of the weak abnormal situation judgment of the potential titration technology is improved, and the error of the analysis result is reduced. BRIEF DESCRIPTION OF DRAWINGS

[0036] Figure 1 The workflow diagram of the potential titration end point prediction and abnormal value correction method based on deep learning of the embodiments of the present application;

[0037] Figure 2A work flow chart for determining whether the potential is a suspected abnormal potential by the potential titration end point prediction and abnormal value correction method based on deep learning of the embodiment of the present application;

[0038] Figure 3 A work flow chart for predicting the abnormal type of the suspected abnormal potential by the potential titration end point prediction and abnormal value correction method based on deep learning of the embodiment of the present application;

[0039] Figure 4 A work flow chart for predicting the current potential as a weak abnormal potential or a normal potential by the potential titration end point prediction and abnormal value correction method based on deep learning of the embodiment of the present application. DETAILED DESCRIPTION

[0040] In order to make the objects and advantages of the present application clearer, the present application will be further described below with reference to the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and not to limit the present application.

[0041] The preferred embodiments of the present application will be described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present application and not to limit the protection scope of the present application.

[0042] In addition, it should be further explained that, in the description of the present application, unless otherwise explicitly specified and limited, the terms "mounting", "connecting", "connection" should be understood in a broad sense, for example, it can be fixed connection, or detachable connection, or integrally connected; it can be mechanical connection, or electrical connection; it can be directly connected, or indirectly connected through an intermediate medium, or the internal communication of two elements. Those skilled in the art can understand the specific meaning of the above terms in the present application according to the specific circumstances.

[0043] Please refer to Figures 1-4 as shown, Figure 1 A work flow chart for the potential titration end point prediction and abnormal value correction method based on deep learning of the embodiment of the present application; Figure 2 A work flow chart for determining whether the potential is a suspected abnormal potential by the potential titration end point prediction and abnormal value correction method based on deep learning of the embodiment of the present application; Figure 3 A work flow chart for predicting the abnormal type of the suspected abnormal potential by the potential titration end point prediction and abnormal value correction method based on deep learning of the embodiment of the present application; Figure 4 A work flow chart for predicting the current potential as a weak abnormal potential or a normal potential by the potential titration end point prediction and abnormal value correction method based on deep learning of the embodiment of the present application.

[0044] The potential titration end point prediction and abnormal value correction method based on deep learning of the embodiment of the present application comprises:

[0045] Step S1, obtaining potential change data in the potentiometric titration process, basic data of the titrant, and parameters of the sample to be measured;

[0046] Step S2, determining whether the potential is a suspected abnormal potential based on whether the potential change slope exceeds a preset change slope and / or whether the potential change violates the monotonicity of the potential change;

[0047] Step S3, predicting the abnormal type of the suspected abnormal potential based on a comparison result of the mutation amplitude of the suspected abnormal potential and the first and second preset mutation amplitudes;

[0048] Step S4, determining whether the current potential is a weak abnormal potential or a normal potential based on whether the local curvature change rate of the adjacent potential in the weak abnormal potential titration is greater than a preset curvature change rate and a time sequence correlation;

[0049] Step S5, determining to correct the abnormal potential titration by different correction methods based on a comparison result of the duration of the abnormal potential and a preset duration and a comparison result of the deviation degree of the abnormal potential titration abnormal value and a preset deviation degree.

[0050] In the embodiment of the application, the potential change data in the potentiometric titration process includes but is not limited to "real-time potential value, potential change rate and second derivative", the basic data of the titrant includes but is not limited to "titrant concentration, titrant type and titrant addition rate", and the parameters of the sample to be measured include but are not limited to "sample initial volume, sample initial pH value and sample temperature".

[0051] Specifically, in step S2, when determining whether the potential is a suspected abnormal potential, it is determined whether the potential is a suspected abnormal potential based on whether the potential change slope exceeds a preset change slope and whether the potential change violates the monotonicity of the potential change;

[0052] If the potential change slope exceeds the preset change slope or the potential change violates the monotonicity of the potential change, it is determined that the potential is a suspected abnormal potential;

[0053] If the potential change slope does not exceed the preset change slope and the potential change conforms to the monotonicity of the potential change, it is determined that the potential is a normal potential.

[0054] In the embodiment of the application, the potential change slope is determined according to the ratio of the potential difference between two adjacent sampling points to the difference in titrant addition amount, and the calculation method is the difference between the current potential and the adjacent potential divided by the difference between the current titrant addition amount and the adjacent titrant addition amount, for example, the current potential value E i = 250 mV, the titrant addition amount V i= 10.00 mL, adjacent potential value E i+1 = 265 mV, titrant added amount V i+1 = 10.10 mL, the potential change slope is (265 mV-250 mV) / (10.10 mL-10 mL)=150 mV / mL, the preset change slope is determined according to the average potential change slope of several same type of potential titration, for example, the average potential change slopes of two groups of same type of potential titration are 140 mV / mL and 160 mV / mL respectively, and the preset change slope is 150 mV / mL, the monotonicity of the potential change is determined according to the consistency of the potential value change trend after the titrant is added, for example, the potential value is detected once a minute as the titrant continues to be added, and the potential values in three minutes are 450 mV, 420 mV and 390 mV respectively, the potential always keeps a downward trend, and it is determined that the potential change conforms to the monotonicity of the potential change, but the above values are not limited thereto, and a person skilled in the art can also adjust the values according to actual needs.

[0055] The present application determines whether the potential is a suspected abnormal potential by whether the potential change slope exceeds the preset change slope and whether the potential change violates the monotonicity of the potential change, according to the potential change slope exceeding the preset change slope or the potential change violating the monotonicity of the potential change, it is explained that the potential mutation or trend abnormality phenomenon occurs, the potential is accurately determined as a suspected abnormal potential, according to the potential change slope not exceeding the preset change slope and the potential change conforming to the monotonicity of the potential change, it is explained that the potential change rate is normal and the trend is consistent, and the potential is accurately determined as a normal potential, the above content improves the accuracy of the weak abnormality judgment of the potential titration technology, and further reduces the error of the analysis result.

[0056] Specifically, in step S3, when predicting the abnormal type of the suspected abnormal potential, the abnormal type of the suspected abnormal potential is predicted based on the suspected abnormal potential and the comparison result of the mutation amplitude of the suspected abnormal potential and the first preset mutation amplitude and the second preset mutation amplitude.

[0057] If the mutation amplitude of the suspected abnormal potential is greater than the first preset mutation amplitude, it is predicted that the abnormal type of the suspected abnormal potential is a strong abnormal potential.

[0058] If the mutation amplitude of the suspected abnormal potential is less than or equal to the first preset mutation amplitude and greater than the second preset mutation amplitude, it is predicted that the abnormal type of the suspected abnormal potential is a weak abnormal potential.

[0059] If the mutation amplitude of the suspected abnormal potential is less than or equal to the second preset mutation amplitude, it is predicted that the potential does not occur abnormality.

[0060] The mutation amplitude of the suspected abnormal potential in the embodiment of the application is determined according to the absolute value of the potential difference between the suspected abnormal potential and the adjacent normal points, and the calculation method is (the potential value of the suspected abnormal potential minus the sum of the potential values of the previous normal potential and the next normal potential) divided by 2. For example, the previous normal point has a potential value E prev = 300 mV; the suspected abnormal point has a potential value E suspected = 350 mV; and the next normal point has a potential value E next = 305 mV, 350 mV-(305 mV+300 mV) / 2 = 47.5 mV, the first preset mutation amplitude range is 30 mV-50 mV, and the preferred value is 40 mV. The reason for selecting the preferred value is based on the statistical characteristics of a large amount of experimental data, which can ensure the sensitivity to true abnormalities and avoid excessive reaction to normal fluctuations. The second preset mutation amplitude is half of the first preset amplitude. For example, the first preset mutation amplitude is 40 mV, and the second preset mutation amplitude is 40 mV / 2 = 20 mV. However, the above values are not limited thereto, and a person skilled in the art can adjust the values according to actual needs.

[0061] The application predicts the abnormal type of the suspected abnormal potential by comparing the suspected abnormal potential and the mutation amplitude of the suspected abnormal potential with the first preset mutation amplitude and the second preset mutation amplitude. According to the fact that the mutation amplitude of the suspected abnormal potential is greater than the first preset mutation amplitude, it is indicated that the potential value has a significant jump and exceeds the upper limit of the normal fluctuation range, and the abnormal type of the suspected abnormal potential is accurately predicted as a strong abnormal potential. According to the fact that the mutation amplitude of the suspected abnormal potential is less than or equal to the first preset mutation amplitude and greater than the second preset mutation amplitude, it is indicated that the potential value has a perceptible abnormal fluctuation but does not reach a significant jump, and the abnormal type of the suspected abnormal potential is accurately predicted as a weak abnormal potential. According to the fact that the mutation amplitude of the suspected abnormal potential is less than or equal to the second preset mutation amplitude, it is indicated that the potential fluctuation is within the normal noise range or only has an inherent instrument error, and it is accurately predicted that the potential does not have an abnormality. The above content improves the accuracy of the weak abnormality judgment of the potential titration technology, and further reduces the error of the analysis result.

[0062] Specifically, in step S4, when the current potential is predicted to be a weak abnormal potential or a normal potential, whether the local curvature change rate of the adjacent potential based on the weak abnormal potential titration is greater than a preset curvature change rate and a time sequence correlation is predicted to be a weak abnormal potential or a normal potential.

[0063] If the local curvature change rate of the adjacent potential based on the weak abnormal potential titration is greater than the preset curvature change rate and the time sequence correlation of the weak abnormal potential is less than the preset time sequence correlation, the current potential is predicted to be a weak abnormal potential.

[0064] If the local curvature change rate of adjacent potential of the weak abnormal potential titration is less than or equal to a preset curvature change rate and the time sequence correlation of the weak abnormal potential is greater than or equal to a preset time sequence correlation, the current potential is predicted as a normal potential.

[0065] The local curvature change rate of adjacent potential of the weak abnormal potential titration in the embodiment of the present application is determined according to the first derivative and the second derivative of potential with respect to time, and the calculation method is that the absolute value of the second derivative of potential with respect to time is divided by the square root of two-thirds of (1 plus the square of the first derivative of potential with respect to time), for example, the first derivative of potential with respect to time is 5 mV / s, and the second derivative of potential with respect to time is 3 mV / s 2 , the curvature change rate is 0.02, the preset curvature change rate is determined according to the average local curvature change rate of adjacent potential of the same type of weak abnormal potential titration, for example, the local curvature change rates of adjacent potential of two groups of the same type of weak abnormal potential titration are 0.03 and 0.04, and the preset curvature change rate is 0.035, but the above-mentioned values are not limited thereto, and the person skilled in the art can also adjust the values according to the actual needs.

[0066] The time sequence correlation in the embodiment of the present application is determined according to the dynamic time warping distance, for example, the current potential sequence is [300, 305], and the normal positioning sequence is [295, 300], a 2*2 matrix is obtained: D=[|300-295||305-295||300-300||305-300|]= [5 10 5], the shortest path is (1, 1)→(1, 2)→(2, 2), the dynamic warping distance is 10, and the time sequence correlation is 0.09 after normalization processing 1 / (1+10) of the dynamic warping distance, the preset time sequence correlation range is set to 0.07-0.09, and the preferred value is 0.08, the reason for selecting the preferred value is that the preferred value is too small, which may miss the early slight abnormality, and the preferred value is too large, which may pass the critical abnormality, but the above-mentioned values are not limited thereto, and the person skilled in the art can also adjust the values according to the actual needs.

[0067] The present application predicts the current potential as a weak abnormal potential or a normal potential by whether the local curvature change rate of the adjacent potential of the weak abnormal potential titration is greater than a preset curvature change rate and a timing correlation, according to the local curvature change rate of the adjacent potential of the weak abnormal potential titration being greater than the preset curvature change rate and the timing correlation of the weak abnormal potential being less than a preset timing correlation, it is indicated that the local bending degree of the potential curve exceeds the normal range, and there is a regularity difference between the current fluctuation mode and the historical normal sequence, accurately predicting the current potential as a weak abnormal potential, according to the local curvature change rate of the adjacent potential of the weak abnormal potential titration being less than or equal to the preset curvature change rate and the timing correlation of the weak abnormal potential being greater than or equal to the preset timing correlation, it is indicated that the potential curve is smooth and the fluctuation mode is highly similar to the historical normal sequence, accurately predicting the current potential as a normal potential, through the above content, the accuracy of the weak abnormal situation judgment of the potential titration technology is improved, and the error of the analysis result is reduced.

[0068] Specifically, in step S5, when it is determined to correct the abnormal potential titration by different correction methods, it is determined to correct the abnormal potential titration by different correction methods based on whether the duration of the abnormal potential is greater than a preset duration and the comparison result of the deviation degree of the abnormal value of the abnormal potential titration and a preset deviation degree;

[0069] If the duration of the abnormal potential is less than the preset duration and the deviation degree of the abnormal value is less than the preset deviation degree, it is determined that the abnormal potential titration does not need to be corrected;

[0070] If the duration of the abnormal potential is less than the preset duration and the deviation degree of the abnormal value of the abnormal potential titration is greater than or equal to the preset deviation degree, it is determined to correct the abnormal potential titration by a K-nearest neighbor algorithm;

[0071] If the duration of the abnormal potential is greater than or equal to the preset duration and the deviation degree of the abnormal value of the abnormal potential titration is less than the preset deviation degree, it is determined to correct the abnormal potential titration by a window sliding method;

[0072] If the duration of the abnormal potential is greater than or equal to the preset duration and the deviation degree of the abnormal value of the abnormal potential titration is greater than or equal to the preset deviation degree, it is determined to re-prepare the potential titration.

[0073] The preset duration range in the embodiment of the application is set to 3s-7s, and the preferred value is 5s. The reason for selecting the preferred value is that a too small preferred value will cause the system to frequently trigger the correction mechanism, and a too large preferred value can cause the subsequent data to continuously deviate from the ideal curve, affecting the accuracy of the titration end point. The deviation degree of the abnormal value is determined according to the absolute value of the potential difference between the abnormal value and the ideal titration curve at the corresponding point. The calculation method is the difference between the actual measured potential and the theoretical potential of the ideal titration curve at the point divided by the theoretical potential of the ideal titration curve at the point. For example, the actual measured potential is 325mV, and the theoretical potential of the ideal titration curve at the point is 300mV, and the abnormal deviation degree is (325mV-300mV) / 300mV=0.08. However, the above-mentioned value is not limited thereto, and a person skilled in the art can also adjust the value according to actual needs.

[0074] In the embodiment of the application, the K-nearest neighbor algorithm is based on the interpolation or correction of normal data points around the abnormal point. For example, the theoretical potential of the ideal titration curve at t=5 is 300mV, and the actual measured value is 325mV (deviation degree 8%, greater than the preset threshold such as 5%). However, the abnormal duration is only 1s (less than the preset duration 5s), K=3 is selected, and the normal points before and after t=5s are taken: t=4s (298mV), t=6s (302mV), t=7s (301mV), and the correction value is calculated as (298mV+302mV+301mV) / 3=300.33mV, that is, the average value of the neighboring points is used to replace the abnormal value 325mV, thereby reducing the deviation degree. The method of window sliding analyzes the historical data trend in the sliding window to correct persistent abnormalities. For example, the window size N is set to 5, and the potential values of the current abnormal single t=10s and the previous four points are taken: 302mV, 303mV, 304mV, 305mV, 306mV, and the window average value is (302mV+303mV+304mV+305mV+306mV) / 5=304mV. If the ideal value at t=10s is 305mV, and the actual measured value is 306mV, then the average value 304mV is used to correct the abnormal value. However, the above-mentioned value is not limited thereto, and a person skilled in the art can also adjust the value according to actual needs.

[0075] The present application corrects the abnormal potential titration by different correction methods according to whether the duration of the abnormal potential is greater than the preset duration and the ratio of the deviation degree of the abnormal value of the abnormal potential titration to the preset deviation degree, according to the duration of the abnormal potential being less than the preset duration and the deviation degree of the abnormal value being less than the preset deviation degree, it is determined that the phenomenon of normal noise which is slightly fluctuated and can be self-recovered, and it is accurately determined that the abnormal potential titration does not need to be corrected, according to the duration of the abnormal potential being less than the preset duration and the deviation degree of the abnormal value of the abnormal potential titration being greater than or equal to the preset deviation degree, it is determined that the phenomenon of isolated abnormality which is sudden strong interference but not sustained, and it is accurately determined that the abnormal potential titration is corrected by K-nearest neighbor algorithm, according to the duration of the abnormal potential being greater than or equal to the preset duration and the deviation degree of the abnormal value of the abnormal potential titration being less than the preset deviation degree, it is determined that the phenomenon of systematic drift which is long-term existing but small amplitude, and it is accurately determined that the abnormal potential titration is corrected by the method of window sliding, according to the duration of the abnormal potential being greater than or equal to the preset duration and the deviation degree of the abnormal value of the abnormal potential titration being greater than or equal to the preset deviation degree, it is determined that the phenomenon of fatal abnormality which is serious and persistent, and it is accurately determined that the potential titration is re-prepared, through the above, the accuracy of the weak abnormal condition judgment of the potential titration technology is improved, and the error of the analysis result is reduced.

[0076] So far, the technical solutions of the present application have been described in combination with the preferred embodiments shown in the drawings, but those skilled in the art can easily understand that the protection scope of the present application is obviously not limited to these specific embodiments. Those skilled in the art can make equivalent changes or replacements to the related technical features without departing from the principles of the present application, and the technical solutions after the changes or replacements will fall within the protection scope of the present application.

Claims

1. A deep learning-based method for predicting the endpoint of potentiometric titration and correcting outliers, characterized in that, include: Step S1: Obtain potential change data, basic data of the titrant, and parameters of the sample to be tested during the potentiometric titration process; Step S2: Based on whether the slope of the potential change exceeds a preset slope and / or whether the potential change violates the monotonicity of the potential change, determine whether the potential is a suspected abnormal potential. Step S3: Based on the comparison results of the mutation amplitude of the suspected abnormal potential with the first preset mutation amplitude and the second preset mutation amplitude, predict the abnormal type of the suspected abnormal potential. Step S4: Based on whether the local curvature change rate of adjacent potentials in the weakly anomalous potential titration is greater than the preset curvature change rate and the temporal correlation, predict whether the current potential is a weakly anomalous potential or a normal potential. Step S5: Based on the comparison results of the duration of the abnormal potential with the preset duration and the comparison results of the deviation of the abnormal value of the abnormal potential titration with the preset deviation, determine the correction method to correct the abnormal potential titration. In step S5, determining to correct the abnormal potential titration using different correction methods based on whether the duration of the abnormal potential is greater than a preset duration and the comparison result of the deviation of the abnormal value of the abnormal potential titration with a preset deviation includes: If the duration of the abnormal potential is less than the preset duration and the deviation of the abnormal value is less than the preset deviation, it is determined that no correction is needed for the abnormal potential titration. If the duration of the abnormal potential is less than the preset duration and the deviation of the abnormal value of the abnormal potential titration is greater than or equal to the preset deviation, it is determined that the abnormal potential titration is corrected using the K-nearest neighbor algorithm. If the duration of the abnormal potential is greater than or equal to the preset duration and the deviation of the abnormal value of the abnormal potential titration is less than the preset deviation, it is determined that the abnormal potential titration will be corrected by the window sliding method. If the duration of the abnormal potential is greater than or equal to a preset duration and the deviation of the abnormal value of the abnormal potential titration is greater than or equal to a preset deviation, it is determined that the potential titration should be prepared again.

2. The deep learning-based potentiometric titration endpoint prediction and outlier correction method according to claim 1, characterized in that, In step S2, determining whether the potential is a suspected abnormal potential based on whether the slope of the potential change exceeds a preset slope and whether the potential change violates the monotonicity of the potential change includes: If the slope of the potential change exceeds a preset slope or the potential change violates the monotonicity of the potential change, the potential is determined to be a suspected abnormal potential. If the slope of the potential change does not exceed the preset slope and the potential change conforms to the monotonicity of the potential change, the potential is determined to be a normal potential.

3. The deep learning-based potentiometric titration endpoint prediction and outlier correction method according to claim 2, characterized in that, The potential change slope is determined based on the ratio of the potential difference between two adjacent sampling points to the difference in the amount of titrant added. The preset change slope is determined based on the average potential change slope during titration of several potentials of the same type. The monotonicity of the potential change is determined based on the trend of potential value change after the addition of titrant.

4. The deep learning-based potentiometric titration endpoint prediction and outlier correction method according to claim 3, characterized in that, In step S3, based on the comparison results of the suspected abnormal potential and the abrupt change amplitude of the suspected abnormal potential with the first preset abrupt change amplitude and the second preset abrupt change amplitude, the abnormality type of the suspected abnormal potential is predicted to include: If the abrupt change amplitude of the suspected abnormal potential is greater than the first preset abrupt change amplitude, the abnormality type of the suspected abnormal potential is predicted to be a strong abnormal potential. If the abrupt change amplitude of the suspected abnormal potential is less than or equal to the first preset abrupt change amplitude and greater than the second preset abrupt change amplitude, the abnormality type of the suspected abnormal potential is predicted to be a weak abnormal potential. If the amplitude of the suspected abnormal potential change is less than or equal to the second preset amplitude, it is predicted that the potential has not become abnormal.

5. The deep learning-based potentiometric titration endpoint prediction and outlier correction method according to claim 4, characterized in that, The abrupt change amplitude of the suspected abnormal potential is determined based on the absolute value of the potential difference between the suspected abnormal potential and the adjacent normal point. The first preset abrupt change amplitude is determined based on the maximum amplitude of normal fluctuations in historical experimental data. The second preset abrupt change amplitude is half of the first preset amplitude.

6. The deep learning-based potentiometric titration endpoint prediction and outlier correction method according to claim 5, characterized in that, In step S4, whether the local curvature change rate of adjacent potentials titrated based on the weak anomalous potential is greater than the preset curvature change rate and the temporal correlation predicts whether the current potential is a weak anomalous potential or a normal potential includes: If the rate of change of local curvature of adjacent potentials in the weak anomalous potential titration is greater than the preset rate of change of curvature and the temporal correlation of the weak anomalous potential is less than the preset temporal correlation, the current potential is predicted to be a weak anomalous potential. If the local curvature change rate of adjacent potentials in the weakly anomalous potential titration is less than or equal to a preset curvature change rate and the temporal correlation of the weakly anomalous potential is greater than or equal to a preset temporal correlation, the current potential is predicted to be a normal potential.

7. The deep learning-based potentiometric titration endpoint prediction and outlier correction method according to claim 6, characterized in that, The rate of change of local curvature of adjacent potentials in the weak anomalous potential titration is determined based on the first and second derivatives of the potential with respect to time, and the preset rate of change of curvature is determined by the average rate of change of local curvature of adjacent potentials in the same type of weak anomalous potential titration.

8. The deep learning-based potentiometric titration endpoint prediction and outlier correction method according to claim 7, characterized in that, The temporal correlation is determined based on the dynamic time warping distance, and the preset temporal correlation is determined based on the temporal correlation distribution of all sequence pairs in historical normal data.

9. The deep learning-based potentiometric titration endpoint prediction and outlier correction method according to claim 1, characterized in that, The preset duration is determined based on the historical average duration of the abnormal potential under standard conditions, the deviation degree of the abnormal value is determined based on the absolute value of the potential difference between the abnormal value and the ideal titration curve at the corresponding point, and the preset deviation degree is determined based on the average deviation degree under several standard conditions of the same type of potential titration.

Images