Online measurement method and system for polyvinyl alcohol alcoholysis degree distribution

Abstract

The present invention relates to the field of online analysis technology, specifically to an online measurement method and system for the alcoholysis degree distribution of polyvinyl alcohol, comprising the following steps: obtaining spectral data of a polyvinyl alcohol solution, extracting characteristic absorption bands during the conversion of ester groups to hydroxyl groups, recording the absorption intensity within the band to construct a sequence, extracting peak positions, calculating intensity differences and marking directions, identifying fluctuation intervals and calculating angular velocity trends, combining the disturbance amplitude with the component overlap rate, screening abnormalities and generating an unstable segment set. The present invention achieves quantitative analysis of alcoholysis degree conversion characteristics by accurately identifying the trend of intensity difference changes in the selection of characteristic absorption bands in spectral data and the intensity difference change trend, thereby improving the accuracy of polyvinyl alcohol alcoholysis degree characterization, and providing more accurate trend fluctuation identification capabilities by comparing the angular velocity difference of the absorption intensity change rate, significantly improving the analytical level of alcoholysis degree distribution details, identifying abnormal fluctuation intervals in alcoholysis degree component distribution, and reducing product quality differences caused by alcoholysis degree fluctuations.

Description

Technical Field

[0001] The present invention relates to the field of online analysis technology, and in particular to an online measurement method and system for the alcoholysis degree distribution of polyvinyl alcohol. Background Art

[0002] The field of online analysis technology involves methods and devices for continuous or real-time detection of the composition, structure, or properties of substances during industrial production or material processing. This technology integrates optical, electrical, chemical, or mechanical sensors, and cooperates with automatic sampling, data acquisition, and processing systems to achieve dynamic monitoring of key parameters of raw materials, intermediates, or final products. It is mainly used in industries such as chemical, pharmaceutical, food, and environmental monitoring to improve process control accuracy, product consistency, and production efficiency, and to avoid information lag caused by human sampling delays. Common online analysis methods include infrared spectroscopy, Raman spectroscopy, mass spectrometry, chromatography, conductivity measurement, near-infrared analysis, etc., which have technical advantages such as fast response, strong non-destructiveness, and high integrability.

[0003] The online measurement method for the alcoholysis degree distribution of polyvinyl alcohol (PVA) aims to obtain real-time information on the alcoholysis degree distribution of PVA materials during their production or application. The alcoholysis degree of PVA, or the extent to which vinyl acetate monomers are converted to alcohol groups, is a key indicator of key performance parameters such as solubility, film-forming properties, and adhesion. This method aims to dynamically monitor the distribution of components with varying alcoholysis degrees in PVA without compromising production continuity, providing an accurate basis for process adjustments, quality control, and product classification. This method is applicable to PVA production lines or application scenarios, improving product consistency and targeted application performance.

[0004] Traditional measurement methods rely on real-time measurement of the overall alcoholysis degree level, and do not effectively focus on the fluctuation details and changing trends in the process of converting ester groups to hydroxyl groups in the molecule. As a result, it is impossible to timely capture local abnormal changes and subtle fluctuations in the alcoholysis degree in the actual production environment. There is a lack of in-depth analysis and quantitative analysis methods for the trend of absorption intensity changes, resulting in insufficient sensitivity in trend fluctuation judgments. It is impossible to quickly respond to and accurately identify abnormal distribution of polyvinyl alcohol alcoholysis degree, and it is difficult to discover and accurately locate unstable component distribution in the product in real time. For example, subtle deviations in alcoholysis degree within a certain period of time during the production process cannot be identified in a timely manner, which will cause frequent fluctuations in product quality and affect the overall consistency and subsequent application performance of polyvinyl alcohol products. Summary of the Invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose an online measurement method and system for the alcoholysis degree distribution of polyvinyl alcohol.

[0006] In order to achieve the above object, the present invention adopts the following technical solution: an online measurement method for the alcoholysis degree distribution of polyvinyl alcohol, comprising the following steps:

[0007] S1: Acquire spectral data of polyvinyl alcohol solution, extract characteristic absorption bands during the conversion of ester groups to hydroxyl groups in polyvinyl alcohol molecules, record all absorption intensities within the band in each sampling cycle, and construct a sequence list by wavelength dimension to generate a characteristic absorption segment spectral sequence;

[0008] S2: Based on the characteristic absorption segment spectrum sequence, the intensity difference change direction of adjacent sampling points is marked, and the frame combination segments with opposite change directions are extracted to identify the start and end frame positions of the fluctuation, and generate a set of absorption peak difference fluctuation intervals;

[0009] S3: calling the absorption peak difference fluctuation interval set, calculating the absorption peak intensity change rate vector in each time interval, calculating the difference between the angular velocity change trend in the current segment and the average absorption intensity change rate in the period, and generating segment absorption response shift trend information;

[0010] S4: Call the absorption response deviation trend information of the segment, extract the ratio of the number of distribution overlap frames to the total number of frames according to the distribution interval of the polyvinyl alcohol molecule alcoholysis components in the corresponding time segment, and construct a ratio relationship between the two parameters to generate the disturbance amplitude and component overlap information.

[0011] The present invention has improvements in that the characteristic absorption segment spectrum sequence includes a wavelength sampling point index set, an absorption intensity multi-frame matrix, an absorption channel time label, an initial position of the peak position within the band, and a sampling period sequence mapping value; the absorption peak difference fluctuation interval set includes a response direction change identification sequence, an absorption intensity difference dynamic index group, an inter-peak amplitude jump point set, fluctuation start and end frame mark information, and absorption segment construction parameters; the segment absorption response offset trend information is specifically an angular velocity change symbol structure, a period difference direction vector group, an absorption trend torsion feature identifier, a vector angle change statistic within a time period, and a spectral offset trend level index; the disturbance amplitude and component overlap information includes a spectral standard deviation value set, the number of component distribution overlap frames within the segment, the signal fluctuation and component matching coefficient, the mean spectral intensity change within the frame segment, and an interference amplitude label index.

[0012] The present invention is improved in that the steps of acquiring the characteristic absorption segment spectrum sequence are specifically as follows:

[0013] S111: Acquire spectral data of the polyvinyl alcohol solution, identify the wavelength position and corresponding absorption intensity value in the spectral data frame in each measurement cycle, organize the original band information according to the measurement time sequence, form a time sequence wavelength intensity comparison table, and generate a spectral original sampling data set;

[0014] S112: Based on the original spectral sampling data set, according to the recorded spectral information, the wavelength bands corresponding to the ester carbonyl stretching vibration region and the hydroxyl stretching vibration region in polyvinyl alcohol are screened, the continuous response values ​​of the two types of bands in each time frame are extracted, and a response set corresponding to the band type and the frame sequence is established to obtain the target functional band response information;

[0015] S113: Call the target function band response information, arrange the response set in order in the wavelength dimension according to each type of band response value sequence and classify it according to the measurement time frame, organize the structure into a wavelength intensity sequence form with frame sequence labels, and generate a characteristic absorption segment spectrum sequence.

[0016] The present invention is improved in that the step of obtaining the absorption peak difference fluctuation interval set is specifically as follows:

[0017] S211: Based on the characteristic absorption segment spectrum sequence, according to the spectrum frame data corresponding to each sampling time point, the peak positions of the ester carbonyl absorption peak band and the hydroxyl absorption peak band are respectively located within the frame, the corresponding absorption intensity values ​​are extracted and a mapping relationship between the two is established to generate a functional peak position intensity control group;

[0018] S212: Calculating the difference between the hydroxyl group absorption peak intensity and the ester group absorption peak intensity based on the functional peak position intensity control group, constructing a time series according to the sampling time sequence, arranging the intensity differences corresponding to the time frames into a continuous change structure, calculating and obtaining the response change trend value of each time frame, marking the trend change direction as positive or negative, and obtaining a peak difference trend change identification sequence;

[0019] S213: Based on the peak difference trend change identification sequence, according to the trend change direction between consecutive frames, the trend symbol reversal point is located and extracted, and one frame is extended on both sides of the reversal point to form a change segment combination, which is marked as the response fluctuation start and end frame combination. The corresponding time index is extracted from the combination and numbered to establish a time period list, and an absorption peak difference fluctuation interval set is generated.

[0020] The present invention is improved in that the step of obtaining the segment absorption response deviation trend information is specifically as follows:

[0021] S311: Calling the absorption peak difference fluctuation interval set, obtaining the corresponding three sets of band response data of the start frame, center frame and end frame, extracting the normalized absorption intensity values ​​of the hydroxyl absorption peak and the ester carbonyl absorption peak in each frame, constructing the first-order difference structure with time as the sequence, obtaining the intensity change path composed of three points, and generating the absorption peak composition vector group;

[0022] S312: Based on the absorption peak vector group, the corresponding vectors of the start frame, center frame, and end frame are extracted. The angle between the two vectors is calculated by coordinate space angle calculation and divided by the time interval to obtain the angular velocity corresponding to each absorption intensity change. The angular velocity calculation results in adjacent time segments are used to compare the direction consistency, construct trend evolution data, and calculate the offset trend value of the absorption peak response segment;

[0023] S313: According to the offset trend value of the absorption peak response segment, the increase and decrease changes of each trend value between two adjacent time periods are retrieved, and the frequency of occurrence of the trend offset turning point in the change direction sequence is calculated. The frequency is combined with the angular velocity direction consistency judgment result to perform cross-statistics, establish a distribution state index table between trend change and directional stability, and generate the segment absorption response offset trend information.

[0024] The present invention is improved in that the steps of obtaining the disturbance amplitude and component coincidence information are specifically as follows:

[0025] S411: Recalling the absorption response shift trend information of the segment, extracting characteristic absorption band response values ​​of the sampling frames from the corresponding time segments marked as having significant changes, performing deviation square processing on the normalized intensity of the main absorption peak in each frame, and calculating the average standard deviation value in the segment to obtain a spectral perturbation amplitude parameter set;

[0026] S412: Calling the spectral perturbation amplitude parameter set, extracting the alcoholysis component distribution interval detected in each frame of the polyvinyl alcohol solution according to the sampling index range of each time period, counting the number of frames with component overlap in each frame within the time period, and comparing it with the total number of frames in the section to form a proportional relationship, calculating and obtaining the composite index of each absorption section, sorting and numbering the sections based on the structural sequence of each section, and generating perturbation amplitude and component overlap information;

[0027] S413: According to the disturbance amplitude and component overlap information, the disturbance level intervals are divided after sorting by index value, the corresponding segment numbers and level states are marked, and the distribution characteristics, disturbance amplitude, and frame overlap structure are uniformly encoded as evaluation parameters to obtain the disturbance amplitude and component overlap information.

[0028] The present invention is improved in that the method further comprises the following steps:

[0029] S5: calling the disturbance amplitude and component coincidence information, comparing it with the alcoholysis degree fluctuation benchmark index threshold, screening the segment frame numbers that deviate from the range, and checking whether there are characteristic changes of two consecutive directions of reversal in the segment absorption response offset trend value of the corresponding frame. If both conditions are met, the time period is marked as a fluctuation abnormal interval, and a segment set marked as an unstable alcoholysis distribution is generated;

[0030] The alcoholysis distribution unstable marking segment set specifically refers to the abnormal fluctuation segment index code, the continuous direction reversal frame label, the stability offset label within the time period, the unstable area judgment logic field and the block status identification label.

[0031] The present invention is improved in that the step of obtaining the alcoholysis distribution unstable marker segment set is specifically as follows:

[0032] S511: Recalling the disturbance amplitude and component overlap information, comparing the composite index data of each segment with the set alcoholysis degree fluctuation benchmark index threshold, screening the segment frame indexes where the composite index deviates from the benchmark threshold, establishing a classification label corresponding to the deviation direction, and obtaining a composite index deviation frame number sequence;

[0033] S512: Based on the composite indicator deviation frame number sequence, extract the angular velocity direction change label of the corresponding frame from the segment absorption response offset trend value, continuously traverse the inter-frame trend change identifiers, count the number of adjacent frame pairs with direction reversal, select segments with more than two groups of angular velocity direction reversal segments, mark them with numbers, and generate a trend direction reversal segment set;

[0034] S513: Call the composite indicator deviation frame number sequence and trend direction reversal segment set, extract the intersection of the two and use the segment frame number as the primary key to merge and mark it as an abnormal state, establish segment labels, start and end frame indexes and stability judgment identifiers, and generate an alcoholysis distribution unstable mark segment set.

[0035] A polyvinyl alcohol alcoholysis degree distribution online measurement system, the system comprising:

[0036] The characteristic spectrum recognition module obtains the spectral data of the polyvinyl alcohol solution, extracts the characteristic absorption band during the conversion of ester groups in the polyvinyl alcohol molecule to hydroxyl groups, records all absorption intensities within the band in each sampling cycle, and constructs a sequence list by wavelength dimension to generate a characteristic absorption segment spectrum sequence;

[0037] The fluctuation interval identification module marks the change direction of the intensity difference of adjacent sampling points based on the characteristic absorption segment spectrum sequence, extracts the frame combination segments with opposite change directions, identifies the start and end frame positions of the fluctuation, and generates a set of absorption peak difference fluctuation intervals;

[0038] The shift trend analysis module calls the absorption peak difference fluctuation interval set, calculates the absorption peak intensity change rate vector in each time interval, calculates the difference between the angular velocity change trend in the current segment and the average absorption intensity change rate in the period, and generates the segment absorption response shift trend information;

[0039] The disturbance amplitude analysis module calls the absorption response deviation trend information of the segment, extracts the ratio of the number of distribution overlap frames to the total number of frames based on the distribution interval of the alcoholysis components of the polyvinyl alcohol molecules in the corresponding time segment, and constructs a ratio relationship between the two parameters to generate disturbance amplitude and component overlap information;

[0040] The unstable marking module calls the disturbance amplitude and component overlap information, compares it with the alcoholysis degree fluctuation benchmark index threshold, filters the segment frame numbers that deviate from the range, and checks whether there are characteristic changes of two consecutive directions of reversal in the segment absorption response offset trend value of the corresponding frame. If both conditions are met at the same time, the time period is marked as a fluctuation abnormality interval, and a segment set of alcoholysis distribution unstable markings is generated.

[0041] Compared with the prior art, the advantages and positive effects of the present invention are:

[0042] In the present invention, by selecting characteristic absorption bands in spectral data and accurately identifying the trend of intensity difference changes, quantitative analysis of the conversion characteristics of the alcoholysis degree is achieved, the characterization accuracy of the polyvinyl alcohol alcoholysis degree is improved, and a complete time series change feature is formed. By comparing the angular velocity difference of the absorption intensity change rate, the sensitivity to capturing weak change trends is enhanced, a more accurate trend fluctuation identification capability is provided, and the level of analysis of the details of the alcoholysis degree distribution is significantly improved. By carefully calculating the relationship between the disturbance amplitude and the component overlap, the abnormal fluctuation range of the alcoholysis degree component distribution can be identified, the positioning accuracy of the fluctuation anomaly is enhanced, the production process adjustment is made more targeted, the difficulty of quality control caused by alcoholysis degree fluctuations is reduced, the stability of the consistency and application performance of polyvinyl alcohol products is improved, an accurate real-time basis is provided for product classification, and the product quality differences caused by alcoholysis degree fluctuations are reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 is a flow chart of the method of the present invention;

[0044] Figure 2 A flow chart for obtaining a characteristic absorption band spectrum sequence for the present invention;

[0045] Figure 3 A flow chart for obtaining a set of absorption peak difference fluctuation intervals according to the present invention;

[0046] Figure 4 A flow chart for obtaining the section absorption response deviation trend information of the present invention;

[0047] Figure 5 A flow chart of obtaining disturbance amplitude and component coincidence information in the present invention;

[0048] Figure 6 The present invention provides a flow chart for obtaining a set of unstable marker segments for alcoholysis distribution. DETAILED DESCRIPTION

[0049] 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.

[0050] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating directions or positional relationships, are based on the directions or positional relationships shown in the accompanying drawings and are only for the convenience of describing the present invention and simplifying the description. They do not indicate or imply that the devices or elements referred to must have a specific direction, be constructed and operate in a specific direction, and therefore should not be understood as limiting the present invention. In addition, in the description of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0051] See also Figure 1 The present invention provides a technical solution: an online measurement method for the alcoholysis degree distribution of polyvinyl alcohol, comprising the following steps:

[0052] S1: Acquire spectral data of polyvinyl alcohol solution, extract characteristic absorption bands during the conversion of ester groups to hydroxyl groups in polyvinyl alcohol molecules, select the ester carbonyl stretching vibration band and the hydroxyl stretching vibration band as comparison targets, record all absorption intensities within the band range in each sampling cycle, and construct a sequence list by wavelength dimension to generate a characteristic absorption segment spectral sequence;

[0053] S2: Based on the characteristic absorption segment spectrum sequence, the ester carbonyl absorption peak and the hydroxyl absorption peak are extracted at each time sampling point, the absorption intensity difference between the two is calculated, and the difference is arranged in chronological order to construct a response change sequence. The change direction of the intensity difference between adjacent sampling points is marked, and the frame combination segments with opposite change directions are extracted. The start and end frame positions of the fluctuation are identified to generate a set of absorption peak difference fluctuation intervals;

[0054] S3: Call the absorption peak difference fluctuation interval set, calculate the absorption peak intensity change rate vector in each time interval based on the marked start frame, center frame and end frame, calculate the corresponding change angular velocity through the angle formed by the three points, compare the directional consistency of the angular velocity in the previous and next cycles, and calculate the difference between the angular velocity change trend in the current segment and the average absorption intensity change rate in the cycle to generate the segment absorption response offset trend information;

[0055] Angular velocity refers to the rate of change of the angle formed by the change in the wavelength-intensity two-dimensional coordinate system. It is often used to measure the time sensitivity of absorption changes during the reaction process, indicating the speed of fluctuation of the absorption change direction per unit time.

[0056] S4: Call the segment absorption response offset trend information, analyze the standard deviation of the spectral signal in all sampling frames according to the segment marked as having a prominent change, and use it as a characterization index of the spectral perturbation amplitude. According to the distribution interval of the alcoholysis components of the polyvinyl alcohol molecules in the corresponding time segment, extract the ratio of the number of distribution overlap frames to the total number of frames, and establish a ratio relationship between the two parameters to generate the perturbation amplitude and component overlap information;

[0057] The standard deviation of the spectral signal refers to the statistical dispersion of the absorption intensity of the characteristic band and is widely used in stability assessment in infrared online analysis. The coincidence ratio index represents the relationship between the spectral stability and the consistency of the component distribution within the multi-component reaction range.

[0058] S5: Call the disturbance amplitude and component overlap information, compare it with the alcoholysis degree fluctuation benchmark index threshold, screen the segment frame numbers that deviate from the range, and check whether there are characteristic changes of two consecutive direction reversals in the segment absorption response offset trend value of the corresponding frame. If both conditions are met at the same time, mark the time period as a fluctuation abnormal interval and generate a segment set marked with unstable alcoholysis distribution;

[0059] The alcoholysis degree fluctuation benchmark index threshold is a preset system stability judgment standard, which is usually derived from the segment distribution and spectral fluctuation experience range under normal production conditions;

[0060] The characteristic absorption segment spectrum sequence includes a wavelength sampling point index set, an absorption intensity multi-frame matrix, an absorption channel time label, an initial peak position within the band, and a sampling period sequence mapping value. The absorption peak difference fluctuation interval set includes a response direction change identification sequence, an absorption intensity difference dynamic index group, an inter-peak amplitude jump point set, fluctuation start and end frame marking information, and absorption segment construction parameters. The segment absorption response offset trend information specifically includes an angular velocity change symbol structure, a period difference direction vector group, an absorption trend torsion characteristic identifier, a vector angle change statistic within a time period, and a spectrum offset trend level index. The disturbance amplitude and component overlap information includes a spectrum standard deviation value set, the number of component distribution overlap frames within the segment, a signal fluctuation and component matching coefficient, a spectral intensity change mean within a frame segment, and an interference amplitude label index. The alcoholysis distribution unstable mark segment set specifically refers to an abnormal fluctuation segment index code, a continuous direction reversal frame label, a stability offset label within a time period, an unstable area judgment logic field, and a block status identification label.

[0061] See also Figure 2 , the specific steps for obtaining the characteristic absorption band spectrum sequence are:

[0062] S111: Acquire spectral data of the polyvinyl alcohol solution, identify the wavelength position and corresponding absorption intensity value in the spectral data frame in each measurement cycle, organize the original band information according to the measurement time sequence, form a time sequence wavelength intensity comparison table, and generate a spectral original sampling data set;

[0063] To obtain the continuous spectrum signal data of the polyvinyl alcohol solution collected by the infrared detection device, it is necessary to set up a two-dimensional acquisition coordinate system with the sampling time as the horizontal axis and the wavelength as the vertical axis. In each cycle, the measurement frequency is set according to the time course of the polyvinyl alcohol hydrolysis reaction in the actual working conditions. The value is usually set to collection Frame, in each frame, all characteristic absorption values ​​and corresponding wavelength points in the signal band are identified to form a wavelength-intensity pair of a single frame. Then, the continuous frame data are integrated and sorted according to the acquisition cycle sequence number, and the response sequence of each wavelength point in continuous time is constructed. In actual implementation, if the solution reaction continues The acquisition frequency is every One frame, a total of Frame, wavelength range is set to , the wavelength step is , then the number of wavelength points per frame is , need to establish Group dimensional vector group, and build a matrix structure with wavelength points as columns and time as rows based on this group of data. The example is as follows:

[0064] Table 1 Absorption signal frame data matrix diagram

[0065] ;

[0066] As shown in Table 1, the data table records the absorption intensity values ​​of specific wavelengths in different time frames, forming a basic matrix structure that can be used for feature extraction. Subsequently, operations such as band interception and target peak identification can be performed in the matrix structure. The structural data is the original spectral sampling data set. This operation process does not involve threshold setting or weight allocation. It only directly reads the values ​​and sorts and classifies them based on the acquisition results of the detection equipment. Therefore, there is no need to establish additional benchmark parameters. However, the consistency of the wavelength sequence in each frame and the completeness of the data frame must be guaranteed. Otherwise, a stable matrix structure sequence cannot be formed.

[0067] S112: Based on the original spectral sampling data set, according to the recorded spectral information, the wavelength bands corresponding to the ester carbonyl stretching vibration region and the hydroxyl stretching vibration region in polyvinyl alcohol are screened, and the continuous response values ​​of the two types of bands in each time frame are extracted. A response set corresponding to the band type and the frame sequence is established to obtain the target functional band response information;

[0068] According to the complete wavelength response data in each frame recorded in the original spectral sampling data set, two types of bands, the ester carbonyl absorption region and the hydroxyl absorption region in polyvinyl alcohol, were screened. The wavelength range of the former was set to , corresponding to the ester carbonyl stretching vibration absorption, the latter band is , attributed to hydroxyl groups Stretching vibration absorption, in each time frame, the absorption intensity values ​​of all wavelength points in the above band are extracted to form a response sequence, and a double index mapping structure is established between the corresponding band and frame index, such as The sequence of ester carbonyl absorption values ​​in the frame is , the hydroxyl absorption value sequence is , corresponding to multiple wavelength points, construct an index table to record the response structure of the frame. The data structure corresponding to the frame number and band type can be expressed as a triple Finally, all frames will be aggregated to form the target function band response matrix. In this process, for each type of band, it is necessary to determine whether its response value presents an upward, platform or downward trend in the same frame, and record the wavelength point position of the median and maximum absorption value as the basis for subsequent feature extraction. If the deviation of the distribution concentration of the response value of a certain band exceeds , that is, the difference between the maximum value and the median exceeds , then the band is marked as a fluctuating state in the current frame, otherwise it is a stable state. The absorption intensity difference is used as a reference value to identify abnormal fluctuations. The measured peak fluctuation range is set in the reaction interval. In actual measurement, polyvinyl alcohol is The average difference between the maximum and median values ​​during alcoholysis is , the control error is set to , the allowed range is ,Exceed will be considered a severe disturbance.

[0069] S113: Call the target function band response information, arrange the response set in order in the wavelength dimension according to each type of band response value sequence and classify it according to the measurement time frame, organize the structure into a wavelength intensity sequence form with frame sequence labels, and generate a characteristic absorption segment spectrum sequence.

[0070] The band response value sequence extracted from the target function band response matrix is ​​called and first arranged in ascending order by wavelength dimension to ensure that the distribution of response data in each frame on wavelength has sequence consistency. Then, the wavelength-intensity pairs are merged into a sequential form frame by frame in sequence to form a three-dimensional matrix structure. The frame sequence is one dimension, the wavelength is two dimensions, and the response value is a three-dimensional variable. A frame sequence-wavelength-response structure is constructed. Each frame forms a column of wavelength response vectors. After all frames are connected, a response surface on the time axis is formed. If the wavelength sampling point is , the sampling frame number is , then the size of the constructed matrix is , sequence style such as , forming a complete structural feature sequence, in which the response evolution path of the peak wavelength point is extracted. If the response value of a wavelength point is continuously The average increase in the frame exceeds , and the maximum value appears in the last frame, then the point is marked as a characteristic transition point and recorded in the peak transition index table. The threshold value of the response change rate is based on the range of changes in the absorption intensity per second. The measured change mean is expanded three times to obtain , , set to As a conventional fluctuation limit, any value exceeding this value is marked as a transition state. Finally, all frame labels, wavelength points, and normalized absorption values ​​are assembled in a structured sequence to form a frame-labeled wavelength intensity sequence form, which is output as a characteristic absorption segment spectral sequence.

[0071] See also Figure 3 , the specific steps for obtaining the absorption peak difference fluctuation interval set are:

[0072] S211: Based on the characteristic absorption segment spectrum sequence, according to the spectrum frame data corresponding to each sampling time point, the peak positions of the ester carbonyl absorption peak band and the hydroxyl absorption peak band are respectively located within the frame, the corresponding absorption intensity values ​​are extracted and a mapping relationship between the two is established to generate a functional peak position intensity control group;

[0073] To obtain the spectral frame data corresponding to each sampling time point in the characteristic absorption segment spectrum sequence, it is necessary to establish a double mapping table between wavelength index and intensity value for each frame in the sequence. First, according to the set ester carbonyl absorption band interval in the frame, and hydroxyl absorption band Perform the maximum value positioning operation, and use the method of comparing the absorption intensity of each wavelength point to screen out the wavelength point with the largest intensity value in the segment and its corresponding value, which is defined as the absorption peak point in the current frame. For example, if Within the frame, the normalized intensity sequence of the hydroxyl absorption segment is , then its maximum value The corresponding wavelength is , which is the hydroxyl peak point; the normalized intensity sequence of the ester absorption segment is , the maximum value is , the corresponding wavelength is , take it as the ester peak point, then record the two wavelength points and their intensity values ​​corresponding to the frame in the structure mapping table, and establish a frame sequence And the one-to-one mapping structure between the two peaks, that is, , extract and summarize the triplet of all frames to form a complete function peak intensity control group. This operation does not involve weight or threshold calculation, only the maximum value in the frame is identified and extracted, and no feature transformation other than data normalization is required. The final output data structure is a double peak structure corresponding to the one-dimensional frame sequence, formatted as: frame ,in , the control group provides basic input for the next trend calculation.

[0074] S212: Based on the functional peak intensity control group, calculate the difference between the hydroxyl absorption peak intensity and the ester absorption peak intensity, and construct a time series according to the sampling time sequence, arranging the intensity differences corresponding to the time frames into a continuous change structure using the formula:

[0075] ;

[0076] The response change trend value of each time frame is obtained by calculation, and the trend change direction is marked as positive or negative, thereby obtaining a peak difference trend change identification sequence;

[0077] in, For time frame The response change trend value, For time frame The normalized value of the hydroxyl absorption peak intensity is For time frame The normalized value of the ester carbonyl absorption peak intensity is For time frame The normalized value of the hydroxyl absorption peak intensity is For time frame The normalized value of the ester carbonyl absorption peak intensity is For time frame The normalized value of the hydroxyl absorption peak intensity is For time frame The normalized value of the hydroxyl absorption peak intensity is The total number of consecutive sampling frames to construct the normalized trend;

[0078] The response change trend value ΔDt is used to quantify the differential change characteristics of the absorption peaks of the two key functional groups, hydroxyl and ester, within the same time frame, and to assist in determining whether there is a synergistic reaction trend between the two.

[0079] ΔDt is essentially a peak dynamic indicator at the single-point level of the time series. It not only considers the instantaneous absorption intensity difference between the hydroxyl group and the ester group (reflecting the spatial difference), but also considers the trend product of these two absorption peaks from the previous frame to the current frame (reflecting the consistency of the time evolution direction). It belongs to the hybrid local trend function dimensional structure.

[0080] Item 1 is the difference in absorption intensity between the hydroxyl group and the ester group in the current frame, indicating the difference in instantaneous response between the functional groups.

[0081] The second term is the direction consistency weighted term:

[0082] molecular It is the trend consistency product. A value greater than zero indicates the same direction, and a value less than zero indicates opposite fluctuations.

[0083] The denominator is the normalized sum of squares of the trend changes in the hydroxyl intensity sequence (to avoid amplification of short-term anomalies), thereby achieving standardization of trend synergy.

[0084] The larger the value, the greater the difference in the intensity of the two peaks, or the trend change direction is consistent and drastic, indicating that the frame is a typical reactive frame. Approaching 0, the two functional groups do not change significantly or in opposite directions, and the reaction trend is unstable;

[0085] According to each set of data extracted from the functional peak intensity control group, the intensity difference change of each frame in the time series is constructed, and the current frame is called respectively. Same as previous frame Normalized intensity values ​​of hydroxyl and ester absorption peaks , perform the product operation of the increase and decrease directions of the two absorption values, analyze whether the change trends of the two peaks are in the same direction, and then combine the absolute difference in intensity between the two peaks in the current frame to calculate the comprehensive change trend value using the following formula:

[0086] ;

[0087] In this calculation process, the sample data is set as follows: , while is the number of traceback frames, and the normalized intensity value sequence is , then:

[0088] The first intensity difference is ;

[0089] The second molecule is ;

[0090] The denominator of the second term is ;

[0091] So the whole formula is:

[0092] ;

[0093] The value is marked with positive and negative change directions. If the trend direction of the current frame is the same as that of the previous frame, it is recorded as "positive change direction". If it is opposite, it is marked as "negative change direction". The marked structure is filed into the trend change identification sequence. Each element in the sequence is In this formula, factors such as "weight" and "coefficient" are not used. All values ​​are directly constructed based on the peak value, and the trend judgment is calculated based on the difference direction relationship, without the need to introduce adjustment items.

[0094] S213: Based on the peak difference trend change identification sequence, according to the trend change direction between consecutive frames, the trend symbol reversal point is located and extracted, and one frame is extended on both sides of the reversal point to form a change segment combination, which is marked as the response fluctuation start and end frame combination. The corresponding time index is extracted from the combination and numbered to establish a time period list, and a set of absorption peak difference fluctuation intervals is generated.

[0095] According to the trend mark of each frame in the peak difference trend change mark sequence, the nodes where the trend symbol appears reversed in the two frames before and after the time series are located and extracted. If a frame is marked as a positive change and the next frame is a negative change, the current frame is the reversal point, and one frame is extended on both sides of the reversal point to form a The frame combination segment structure is extracted and marked as a fluctuation segment. The inversion combination structure is repeatedly extracted in multiple frame segments and numbered to form a time index sequence. The frame number range is also marked. For example, , and are uniformly filed in a response fluctuation segment index table, set as (segment number, start frame, end frame), so that subsequent structures can identify and compare trend features in segmented form. Ultimately, all fluctuation segment identifiers are collated and output as a set of absorption peak difference fluctuation intervals. This process does not involve any threshold setting or benchmark judgment. All fluctuation segments are constructed solely based on the number of trend reversal occurrences and frame index calculation results, which is a direct calculation structure extraction.

[0096] See also Figure 4 , the specific steps for obtaining the segment absorption response deviation trend information are as follows:

[0097] S311: Call the absorption peak difference fluctuation interval set to obtain the corresponding three sets of band response data: the start frame, the center frame, and the end frame. Extract the normalized absorption intensity values ​​of the hydroxyl absorption peak and the ester carbonyl absorption peak in each frame, and construct a first-order difference structure with time as the sequence to obtain the intensity change path composed of three points, and generate the absorption peak composition vector group;

[0098] To call each time period index recorded in the absorption peak difference fluctuation interval set, it is necessary to first clarify the specific numbers of the start frame, intermediate frame and end frame marked in each time period, and extract the complete wavelength response data under the corresponding frame number from the characteristic absorption segment spectrum sequence. In this data, according to the ester carbonyl absorption band Hydroxyl absorption band , respectively find the wavelength point with the maximum normalized intensity in each frame, and extract its intensity value as the absorption peak response value within the frame to form Set a segment number to , whose frame number is , then extract the response values ​​from the three frames to construct a data sequence, for example, the hydroxyl absorption intensity of the 22nd frame , ester absorption intensity ; Hydroxyl absorption intensity in frame 23 , ester absorption intensity ; Hydroxyl absorption intensity in frame 24 , ester absorption intensity Then, the time frame sequence is used as the vector construction benchmark to calculate the first-order difference between the absorption values ​​of each molecular group. For example, the hydroxyl vector is , the differences are and ; The ester carbonyl vector is , the difference is and The change vectors from the start frame to the center frame and from the center frame to the end frame are constructed using two sets of difference sequences to form , , that is, the intensity change path composed of three points, and recorded in the absorption peak vector group, providing a data basis for the next step of angular velocity and trend direction analysis.

[0099] S312: Based on the absorption peaks, the vector groups corresponding to the start frame, center frame, and end frame are extracted. The angle between the two vectors is calculated by the coordinate space angle operation and divided by the time interval to obtain the angular velocity corresponding to each absorption intensity change. The angular velocity calculation results in adjacent time segments are used to compare the direction consistency and construct the trend evolution data using the formula:

[0100] ;

[0101] Calculate and obtain the shift trend value of the absorption peak response section;

[0102] in, For the The shift trend value of the absorption peak response segment, For the In the absorption peak response section, the absorption intensity change vector from the start frame to the center frame is For the In the absorption peak response section, the absorption intensity change vector from the center frame to the end frame is formed. For the The dot product of the two absorption intensity change vectors in the segment, For the Vectors in segments The module length, For the Vectors in segments The module length, For the In an absorption peak response segment, the average value of the angles between all absorption intensity change vectors composed of the time frames contained in the segment, For the In an absorption peak response segment, the standard deviation of the angles between all absorption intensity change vectors composed of the time frames contained in the segment;

[0103] The offset trend value Ψz is a segment-level structural response quantity, which is used to quantify the degree of angle bending of a three-frame absorption peak intensity path in the vector space and measure the trend consistency and structural stability of the intensity change path.

[0104] Ψz is a typical angle-dimensional characteristic, reflecting the nonlinear offset between the two vector directions, while taking into account the difference between local trend changes and the overall stability of the segment.

[0105] It is the before-after change vector (temporal gradient of absorption intensity) composed of three frame segments;

[0106] It is calculated as the angle, reflecting the degree of trend reversal;

[0107] is the mean and standard deviation of the vector angles of all three frames in the segment, used to normalize the angle change;

[0108] The numerator reflects the degree of trend deviation, and the denominator is used to suppress the weight of the large fluctuation segment;

[0109] A large Ψz indicates that there is a trend reversal or mutation in the absorption path of the current segment; a Ψz approaching 0 indicates a stable trend change and a small angle deviation, which is a sign of a stable system response.

[0110] According to the absorption peak, the three-point data structure in the vector group is formed, and the cosine angle formula is used to calculate the vector in the two-dimensional vector space. and Calculate the angle to determine the degree of trend bending and divide it by the time interval to obtain the angular velocity value. , For example, first calculate the inner product as , the module lengths are , , and then we get:

[0111] ;

[0112] This value indicates the The absorption response trend deviation of the segment is then divided into all trend values ​​in different segments. The data are numbered and recorded in chronological order to form a trend sequence structure. The benefit of this formula is that by simultaneously incorporating the inner product, modulus, and normalized angle dispersion parameters, the trend shift expression does not rely on the rate of change of a single frame, but instead reflects the structural response relationship between the combination of multiple directional changes and their stable baseline deviation, effectively characterizing the structural shift performance of absorbing reversals within the segment.

[0113] S313: According to the offset trend value of the absorption peak response segment, the increase and decrease changes of each trend value between two adjacent time periods are retrieved, and the frequency of occurrence of the trend offset turning point in the change direction sequence is calculated. The frequency is combined with the angular velocity direction consistency judgment result to perform cross-statistics, establish a distribution state index table between trend change and directional stability, and generate the segment absorption response offset trend information.

[0114] According to the absorption response offset trend value sequence, between two adjacent time periods, the trend value Extract and record the direction of increase and decrease, mark the positive increase as "up", mark the negative decrease as "down", and mark all the sign reversal nodes of "up→down" or "down→up" with frame sequence index, and count their cumulative frequency in the whole sequence. For example, if there are 14 direction reversal points in 45 segments, the frequency is At the same time, the number of positive and negative switching of the angular velocity direction change trend of each segment is extracted. For example, if the direction change mode of segment z in the frame sequence is "positive-positive-negative-negative-positive", the direction switching is counted twice. Combined with the above reversal frequency information, the trend-direction stability cross structure is established by comparing the intersection positions of the two in adjacent segments, and finally a unified record structure index is recorded. , forming a complete trend and direction distribution mapping information, and outputting the segment absorption response offset trend information. No additional thresholds, weights, or coefficients are introduced in this process. All structure construction is based solely on the relationship between trend values ​​and direction sequences, without the need to set parameters or adjust items, ensuring that the data results are derived from the characteristics of the response sequence itself.

[0115] See also Figure 5 , the specific steps for obtaining the disturbance amplitude and component coincidence information are:

[0116] S411: Recall the segment absorption response deviation trend information, extract the characteristic absorption band response value of the sampling frame from the corresponding time segment marked as a prominent change, perform deviation square processing on the normalized intensity of the main absorption peak in each frame, and calculate the average standard deviation value in the segment to obtain the spectral perturbation amplitude parameter set;

[0117] Call the time segment marked as having a prominent change in the segment absorption response deviation trend value, first confirm the frame range index number of each marked segment, and extract the complete wavelength response structure in the corresponding frame in the characteristic absorption segment spectrum sequence, locate the maximum normalized absorption intensity of the ester carbonyl absorption peak and the hydroxyl absorption peak in each frame, record its value as the main absorption peak, and establish a sequence set ,in is the segment number, is the frame index. Then calculate the mean difference between the main peak value in each frame and the main peak value of all frames in the segment, that is, Normalized intensity value of the absorption peak within the frame , first find the average value of the segment ,in is the total number of frames in the segment, and then the square of the difference is performed on each frame , and take the average of all the results in the whole segment and then take the root to get the standard deviation of the segment. The sampling frame is set in the segment Frame, the absorption peak normalized intensity values ​​are , calculate its mean , the square deviations of each frame are , and the sum is , the standard deviation is , record this value as the spectrum perturbation amplitude parameter of the current segment, and perform the above operations on all segments in turn to form a spectrum perturbation amplitude parameter set.

[0118] S412: Call the spectrum perturbation amplitude parameter set, extract the alcoholysis component distribution interval of each frame detected in the polyvinyl alcohol solution according to the sampling index range of each time period, count the number of frames with overlapping components in each frame within the time segment, and compare it with the total number of frames in the segment to form a proportional relationship, using the formula:

[0119] ;

[0120] The composite index of each absorption segment is obtained by calculation, and the structural sequence of each segment is combined to sort and number them, generating the perturbation amplitude and component overlap information;

[0121] in, Indicates the The composite index of the absorption segment, Indicates the In the section The normalized intensity value of the absorption peak of the frame, represents the mean absorption intensity of all sampling frames in the segment, Indicates the The number of frames with overlapping components in a segment, Indicates the The total number of sampling frames in the segment, Indicates the Theoretical overlap ratio of components within each segment;

[0122] The composite index Φk is used to measure the comprehensive performance between the spectral perturbation amplitude of a certain absorption segment and the overlap between the distribution of alcoholysis components. It is a composite stability evaluation index with mixed dimensions.

[0123] This index integrates,the discreteness of the spectral response reflecting the perturbation intensity,the gap between the measured coincidence rate and the theoretical stable coincidence rate (reflecting the consistency of the components);

[0124] Φk serves as a coordination index between spectral stability and response consistency in the system;

[0125] The numerator is the standard deviation (fluctuation amplitude) of the main absorption peak intensity within the segment;

[0126] The denominator is the deviation from component consistency:

[0127] is the actual component overlap rate;

[0128] The ideal overlap ratio set theoretically (usually derived from experimental conditions);

[0129] The larger the difference, the more serious the distribution deviation.

[0130] The larger Φk is, the more severe the fluctuation is, and the component overlap deviates from expectations, reflecting the poor stability of the section. Φk is close to 0: it indicates that the disturbance is small and the components are well matched, which represents the ideal state of the system;

[0131] According to the spectrum perturbation amplitude parameter, each segment is concentrated The time frame index range is extracted frame by frame to extract the alcoholysis component distribution information in the polyvinyl alcohol solution and obtain the distribution interval in each frame, such as The alcoholysis components of the frame are , and the previous frame There are overlapping intervals , then the frame is marked as a "component overlap frame", and the numbers of all frames with overlap intervals in the segment are accumulated to get the number of component overlap frames. and the total number of frames Divide to form a ratio. Introduce a segment The corresponding theoretical overlap ratio This value is derived from the consistency ratio of the component interval of polyvinyl alcohol in the experimental stable state under the process setting conditions, with the alcoholysis rate in the stable operating conditions. The overlap ratio of interval fluctuations is used as a reference, and the theoretical overlap ratio is set as . Substitute into the formula:

[0132] ;

[0133] Taking the results in paragraph 1 as an example, let the standard deviation be , segment frame number , overlapping frames , then the ratio is , the denominator is calculated as , the final composite index is , all segments After aggregation, they correspond one-to-one with the segment index structure, are numbered according to time sequence or composite value size to form a sequence, and the output is the disturbance amplitude and component overlap information.

[0134] S413: According to the disturbance amplitude and component overlap information, the disturbance level intervals are divided after sorting by index value, the corresponding segment numbers and level status are marked, and the distribution characteristics, disturbance amplitude, and frame overlap structure are uniformly encoded as evaluation parameters to obtain the disturbance amplitude and component overlap information.

[0135] According to the composite index corresponding to each segment in the disturbance amplitude and component overlap information , sort its values ​​from small to large, divide the disturbance level interval, and set the baseline level threshold as , the corresponding disturbance levels are low, medium and high respectively. Each segment is classified and marked according to the interval of its composite index, such as Then, the associated attributes such as the section number, disturbance level, location, overlap rate, standard deviation, etc. are coded to form a multi-field structure index table. The table format is as follows: , forming a complete feature assessment structure. The output is information on the disturbance amplitude and component overlap, which provides a quantitative reference for downstream stability assessment and abnormal segment screening. This process does not involve the establishment of new thresholds or weights; all operations are completed based on the established indicator structure and existing numerical inputs.

[0136] See also Figure 6 The specific steps for obtaining the alcoholysis distribution unstable marker segment set are as follows:

[0137] S511: Recall the disturbance amplitude and component overlap information, compare the composite index data of each segment with the set alcoholysis degree fluctuation benchmark index threshold, select the segment frame index where the composite index deviates from the benchmark threshold, establish a classification label corresponding to the deviation direction, and obtain the composite index deviation frame number sequence;

[0138] Call the composite index of each segment in the disturbance amplitude and component overlap information As input data, the alcoholysis degree fluctuation benchmark index threshold must be set first to identify whether the current section has a stability offset. The threshold is obtained through empirical evaluation combined with the fluctuation safety interval setting. In experimental statistics, when the composite fluctuation index value of polyvinyl alcohol exceeds When the product performance begins to show problems such as uneven film thickness and viscosity deviation, the alcoholysis degree fluctuation benchmark index threshold is set to This value is used to determine whether the current segment index deviates from the benchmark state. and Compare, if satisfied or , it is marked as "upward bias" or "downward bias" type, otherwise it is considered "stable".

[0139] For example: If the segment number The composite index is , then the difference is , belongs to "upward bias"; if the segment The composite index is , the difference is , belongs to "downward deviation", and finally forms a corresponding set of frame number sequence and deviation direction label, such as , this set structure is the composite indicator deviation frame number sequence, which is used for subsequent trend direction judgment processing.

[0140] S512: Based on the composite indicator deviation frame number sequence, extract the angular velocity direction change label of the corresponding frame from the segment absorption response offset trend value, continuously traverse the inter-frame trend change identifiers, count the number of adjacent frame pairs with direction reversal, select segments with more than two cumulative groups, mark them and number them, and generate a trend direction reversal segment set;

[0141] All segment numbers extracted from the frame number sequence according to the composite index deviation , sequentially retrieve the angular velocity direction change information recorded in the corresponding segment absorption response deviation trend value, compare the direction identification of the current frame and the next frame frame by frame in each deviation segment, record the "positive→negative" or "negative→positive" reversal behavior, and accumulate statistics for the frame pairs with direction reversal. If the number of consecutive direction reversals in the same segment exceeds two groups, it is determined to be a trend reversal segment. For example, if the segment The frame sequence is , the direction signs are ,but 、 There are two reversals, plus If it is "-→+", it is counted as three reversals, meeting the screening criteria greater than 2, and added to the reversal identification segment set. Finally, all the segment numbers that meet the reversal conditions are sorted into a structure set in the format of The output is a set of trend direction reversal segments, which are then cross-matched with the deviation frame number.

[0142] S513: Call the composite indicator deviation frame number sequence and trend direction reversal segment set, extract the intersection of the two and merge and mark them as abnormal states with the segment frame number as the primary key, establish segment labels, start and end frame indexes and stability judgment identifiers, and generate an alcoholysis distribution unstable mark segment set.

[0143] Call all the segment number information extracted from the composite indicator deviation frame number sequence and the trend direction reversal segment set, perform set intersection calculation, perform matching and merging operations on the numbers that appear in both sets, confirm that they are the segments where the stability structure is abnormal, use this intersection structure as the basis for abnormality identification, and record the segment start frame number in each successfully matched segment. , end frame number and original trend offset value , and set the status field to "Exception Mark". The unified output structure table entry is After all matching items are aggregated, the output is a set of segments marked as unstable alcoholysis distributions. This structure serves as the core criterion for determining abnormal system fluctuations. This process is based solely on the combination of existing numerical calculation results and does not involve additional parameter settings or assumed functions, ensuring a closed-loop data source.

[0144] An online measurement system for polyvinyl alcohol alcoholysis degree distribution, the system comprising:

[0145] The characteristic spectrum recognition module obtains the spectral data of the polyvinyl alcohol solution, extracts the characteristic absorption band during the conversion of ester groups in the polyvinyl alcohol molecule to hydroxyl groups, records all absorption intensities within the band in each sampling cycle, and constructs a sequence list by wavelength dimension to generate a characteristic absorption segment spectrum sequence;

[0146] The fluctuation interval recognition module marks the change direction of the intensity difference between adjacent sampling points based on the characteristic absorption segment spectrum sequence, extracts the frame combination segments with opposite change directions, identifies the start and end frame positions of the fluctuation, and generates a set of absorption peak difference fluctuation intervals;

[0147] The shift trend analysis module calls the absorption peak difference fluctuation interval set, calculates the absorption peak intensity change rate vector in each time interval, calculates the difference between the angular velocity change trend in the current segment and the average absorption intensity change rate in the period, and generates the segment absorption response shift trend information;

[0148] The disturbance amplitude analysis module uses the segment absorption response deviation trend information and extracts the ratio of the number of distribution overlap frames to the total number of frames based on the distribution interval of the polyvinyl alcohol molecule alcoholysis components in the corresponding time segment. The ratio relationship between the two parameters is then established to generate the disturbance amplitude and component overlap information.

[0149] The unstable marking module calls the disturbance amplitude and component overlap information, compares it with the alcoholysis degree fluctuation benchmark index threshold, filters the segment frame numbers that deviate from the range, and checks whether there are characteristic changes of two consecutive direction reversals in the segment absorption response offset trend value of the corresponding frame. If both conditions are met at the same time, the time period is marked as a fluctuation abnormality interval, and a segment set of unstable alcoholysis distribution markings is generated.

[0150] The above are merely preferred embodiments of the present invention and do not limit the present invention in any other form. Any technician familiar with the profession may use the technical content disclosed above to change or modify it into an equivalent embodiment with equivalent changes and apply it to other fields. However, any simple modification, equivalent change and modification made to the above embodiment based on the technical essence of the present invention without departing from the content of the technical solution of the present invention shall still fall within the scope of protection of the technical solution of the present invention.

Claims

1. An online measurement method for the alcoholysis degree distribution of polyvinyl alcohol, characterized in that: The following steps are involved: S1: Acquire spectral data of polyvinyl alcohol solution, extract characteristic absorption bands during the conversion of ester groups to hydroxyl groups in polyvinyl alcohol molecules, record all absorption intensities within the band in each sampling cycle, and construct a sequence list by wavelength dimension to generate a characteristic absorption segment spectral sequence; S2: Based on the characteristic absorption segment spectrum sequence, the intensity difference change direction of adjacent sampling points is marked, and the frame combination segments with opposite change directions are extracted to identify the start and end frame positions of the fluctuation, and generate a set of absorption peak difference fluctuation intervals; S3: calling the absorption peak difference fluctuation interval set, calculating the absorption peak intensity change rate vector in each time interval, calculating the difference between the angular velocity change trend in the current segment and the average absorption intensity change rate in the period, and generating segment absorption response shift trend information; S4: calling the absorption response deviation trend information of the segment, extracting the ratio of the number of distribution overlap frames to the total number of frames according to the distribution interval of the alcoholysis components of the polyvinyl alcohol molecules in the corresponding time segment, and constructing a ratio relationship between the two parameters to generate the disturbance amplitude and component overlap information; S5: calling the disturbance amplitude and component coincidence information, comparing it with the alcoholysis degree fluctuation benchmark index threshold, screening the segment frame numbers that deviate from the range, and checking whether there are characteristic changes of two consecutive directions of reversal in the segment absorption response offset trend value of the corresponding frame. If both conditions are met, the time period is marked as a fluctuation abnormal interval, and a segment set marked as an unstable alcoholysis distribution is generated; The alcoholysis distribution unstable marking segment set specifically refers to the abnormal fluctuation segment index code, the continuous direction reversal frame label, the stability offset label within the time period, the unstable area judgment logic field and the block status identification label.

2. The online measurement method for polyvinyl alcohol alcoholysis degree distribution according to claim 1, characterized in that: The characteristic absorption segment spectrum sequence includes a wavelength sampling point index set, an absorption intensity multi-frame matrix, an absorption channel time label, an initial peak position within the band, and a sampling period sequence mapping value. The absorption peak difference fluctuation interval set includes a response direction change identification sequence, an absorption intensity difference dynamic index group, an inter-peak amplitude jump point set, fluctuation start and end frame mark information, and absorption segment construction parameters. The segment absorption response offset trend information is specifically an angular velocity change symbol structure, a period difference direction vector group, an absorption trend torsion characteristic identifier, a vector angle change statistic within a time period, and a spectral offset trend level index. The disturbance amplitude and component overlap information includes a spectral standard deviation value set, the number of component distribution overlap frames within the segment, the signal fluctuation and component matching coefficient, the mean spectral intensity change within the frame segment, and an interference amplitude label index.

3. The online measurement method for the alcoholysis degree distribution of polyvinyl alcohol according to claim 2, wherein: The steps for acquiring the characteristic absorption segment spectrum sequence are specifically as follows: S111: Acquire spectral data of the polyvinyl alcohol solution, identify the wavelength position and corresponding absorption intensity value in the spectral data frame in each measurement cycle, organize the original band information according to the measurement time sequence, form a time sequence wavelength intensity comparison table, and generate a spectral original sampling data set; S112: Based on the original spectral sampling data set, according to the recorded spectral information, the wavelength bands corresponding to the ester carbonyl stretching vibration region and the hydroxyl stretching vibration region in polyvinyl alcohol are screened, the continuous response values ​​of the two types of bands in each time frame are extracted, and a response set corresponding to the band type and the frame sequence is established to obtain the target functional band response information; S113: Call the target function band response information, arrange the response set in order in the wavelength dimension according to each type of band response value sequence and classify it according to the measurement time frame, organize the structure into a wavelength intensity sequence form with frame sequence labels, and generate a characteristic absorption segment spectrum sequence.

4. The online measurement method for polyvinyl alcohol alcoholysis degree distribution according to claim 3, characterized in that: The steps for obtaining the absorption peak difference fluctuation interval set are specifically as follows: S211: Based on the characteristic absorption segment spectrum sequence, according to the spectrum frame data corresponding to each sampling time point, the peak positions of the ester carbonyl absorption peak band and the hydroxyl absorption peak band are respectively located within the frame, the corresponding absorption intensity values ​​are extracted and a mapping relationship between the two is established to generate a functional peak position intensity control group; S212: Calculating the difference between the hydroxyl group absorption peak intensity and the ester group absorption peak intensity based on the functional peak position intensity control group, constructing a time series according to the sampling time sequence, arranging the intensity differences corresponding to the time frames into a continuous change structure, calculating and obtaining the response change trend value of each time frame, marking the trend change direction as positive or negative, and obtaining a peak difference trend change identification sequence; S213: Based on the peak difference trend change identification sequence, according to the trend change direction between consecutive frames, the trend symbol reversal point is located and extracted, and one frame is extended on both sides of the reversal point to form a change segment combination, which is marked as the response fluctuation start and end frame combination. The corresponding time index is extracted from the combination and numbered to establish a time period list, and an absorption peak difference fluctuation interval set is generated.

5. The online measurement method for polyvinyl alcohol alcoholysis degree distribution according to claim 4, characterized in that: The steps for obtaining the segment absorption response deviation trend information are specifically as follows: S311: Calling the absorption peak difference fluctuation interval set, obtaining the corresponding three sets of band response data of the start frame, center frame and end frame, extracting the normalized absorption intensity values ​​of the hydroxyl absorption peak and the ester carbonyl absorption peak in each frame, constructing the first-order difference structure with time as the sequence, obtaining the intensity change path composed of three points, and generating the absorption peak composition vector group; S312: Based on the absorption peak vector group, the corresponding vectors of the start frame, center frame, and end frame are extracted. The angle between the two vectors is calculated by coordinate space angle calculation and divided by the time interval to obtain the angular velocity corresponding to each absorption intensity change. The angular velocity calculation results in adjacent time segments are used to compare the direction consistency, construct trend evolution data, and calculate the offset trend value of the absorption peak response segment; S313: According to the offset trend value of the absorption peak response segment, the increase and decrease changes of each trend value between two adjacent time periods are retrieved, and the frequency of occurrence of the trend offset turning point in the change direction sequence is calculated. The frequency is combined with the angular velocity direction consistency judgment result to perform cross-statistics, establish a distribution state index table between trend change and directional stability, and generate the segment absorption response offset trend information.

6. The online measurement method for polyvinyl alcohol alcoholysis degree distribution according to claim 5, characterized in that: The steps for obtaining the disturbance amplitude and component coincidence information are specifically as follows: S411: Recalling the absorption response shift trend information of the segment, extracting characteristic absorption band response values ​​of the sampling frames from the corresponding time segments marked as having significant changes, performing deviation square processing on the normalized intensity of the main absorption peak in each frame, and calculating the average standard deviation value in the segment to obtain a spectral perturbation amplitude parameter set; S412: Calling the spectral perturbation amplitude parameter set, extracting the alcoholysis component distribution interval detected in each frame of the polyvinyl alcohol solution according to the sampling index range of each time period, counting the number of frames with component overlap in each frame within the time period, and comparing it with the total number of frames in the section to form a proportional relationship, calculating and obtaining the composite index of each absorption section, sorting and numbering the sections based on the structural sequence of each section, and generating perturbation amplitude and component overlap information; S413: According to the disturbance amplitude and component overlap information, the disturbance level intervals are divided after sorting by index value, the corresponding segment numbers and level states are marked, and the distribution characteristics, disturbance amplitude, and frame overlap structure are uniformly encoded as evaluation parameters to obtain the disturbance amplitude and component overlap information.

7. The online measurement method for polyvinyl alcohol alcoholysis degree distribution according to claim 6, characterized in that: The steps for obtaining the alcoholysis distribution unstable marker segment set are specifically as follows: S511: Recalling the disturbance amplitude and component overlap information, comparing the composite index data of each segment with the set alcoholysis degree fluctuation benchmark index threshold, screening the segment frame indexes where the composite index deviates from the benchmark threshold, establishing a classification label corresponding to the deviation direction, and obtaining a composite index deviation frame number sequence; S512: Based on the composite indicator deviation frame number sequence, extract the angular velocity direction change label of the corresponding frame from the segment absorption response offset trend value, continuously traverse the inter-frame trend change identifiers, count the number of adjacent frame pairs with direction reversal, select segments with more than two groups of angular velocity direction reversal segments, mark them with numbers, and generate a trend direction reversal segment set; S513: Call the composite indicator deviation frame number sequence and trend direction reversal segment set, extract the intersection of the two and use the segment frame number as the primary key to merge and mark it as an abnormal state, establish segment labels, start and end frame indexes and stability judgment identifiers, and generate an alcoholysis distribution unstable mark segment set.

8. An online measurement system for polyvinyl alcohol alcoholysis degree distribution, characterized in that: The system is used to implement the online measurement method for the polyvinyl alcohol alcoholysis degree distribution according to any one of claims 1 to 7, and the system comprises: The characteristic spectrum recognition module obtains the spectral data of the polyvinyl alcohol solution, extracts the characteristic absorption band during the conversion of ester groups in the polyvinyl alcohol molecule to hydroxyl groups, records all absorption intensities within the band in each sampling cycle, and constructs a sequence list by wavelength dimension to generate a characteristic absorption segment spectrum sequence; The fluctuation interval identification module marks the change direction of the intensity difference of adjacent sampling points based on the characteristic absorption segment spectrum sequence, extracts the frame combination segments with opposite change directions, identifies the start and end frame positions of the fluctuation, and generates a set of absorption peak difference fluctuation intervals; The shift trend analysis module calls the absorption peak difference fluctuation interval set, calculates the absorption peak intensity change rate vector in each time interval, calculates the difference between the angular velocity change trend in the current segment and the average absorption intensity change rate in the period, and generates the segment absorption response shift trend information; The disturbance amplitude analysis module calls the absorption response deviation trend information of the segment, extracts the ratio of the number of distribution overlap frames to the total number of frames based on the distribution interval of the alcoholysis components of the polyvinyl alcohol molecules in the corresponding time segment, and constructs a ratio relationship between the two parameters to generate disturbance amplitude and component overlap information; The unstable marking module calls the disturbance amplitude and component overlap information, compares it with the alcoholysis degree fluctuation benchmark index threshold, filters the segment frame numbers that deviate from the range, and checks whether there are characteristic changes of two consecutive directions of reversal in the segment absorption response offset trend value of the corresponding frame. If both conditions are met at the same time, the time period is marked as a fluctuation abnormality interval, and a segment set of alcoholysis distribution unstable markings is generated.

Images