Method and device for detecting uniformity in bag shaking and subpackaging process of particle products

Abstract

The invention relates to the technical field of granular product production, and discloses a uniformity detection method and device in a bag shaking subpackaging process of granular products, and the method comprises the steps: obtaining a whole-process spectrum of a target granular product mixing stage, and determining a mixing end point of the target granular product according to the whole-process spectrum; collecting a plurality of spectrum samples according to the mixing end point, and establishing a principal component analysis model according to the plurality of spectrum samples; determining a warning line and a control line of Hotelling statistics by using a principal component analysis model and the plurality of spectral samples; and acquiring a real-time spectrum of the target particle product in the bag shaking sub-packaging process, and carrying out uniformity detection on the particle product in the bag shaking sub-packaging process by utilizing the principal component analysis model and the warning line and the control line of the Hotelling statistics. According to the method, the spectrum in the mixing stage is used as a uniformity analysis standard, and the uniformity of the particle product in the bag shaking and subpackaging process is detected in real time, so that the uniformity of the material after bag shaking and mixing before bagging is consistent with that of the qualified material in the mixing stage.

Description

Technical Field

[0001] The present invention relates to the technical field of granular product production, and particularly relates to a method and device for detecting the uniformity during the shaking bag packaging process of granular products. Background Art

[0002] During the production process of granular products (such as calcium carbonate D3 granules), it is necessary to mix intermediates and excipients and package them in the form of shaking bags into finished product packaging bags to obtain finished products. The mixing and packaging process includes two stages: the intermediate and excipient mixing stage and the shaking bag packaging stage. Only after the materials are evenly mixed in the mixing stage can they enter the shaking bag packaging stage. However, during the shaking bag packaging process, due to the different molecular sizes of the intermediates and excipients, the sedimentation times are different, resulting in uneven materials entering the finished product packaging bags, which is different from the preset standard substance for the detection of the mixing end point. And this difference is not known to the staff, making the packaged finished products uneven, and then leading to the unqualified finished products leaving the warehouse, causing various problems in the subsequent stages (such as the whole batch of finished products being recognized as unqualified due to unqualified sampling, which is time-consuming and laborious). If the uneven problem that occurs during the shaking bag packaging process can be detected in time, the problem can be eliminated in the early stage of leaving the warehouse, improving the production and delivery efficiency. Summary of the Invention

[0003] In view of this, the present invention provides a method and device for detecting the uniformity during the shaking bag packaging process of granular products to solve the problem of uneven mixing of materials during the shaking bag packaging process.

[0004] In a first aspect, the present invention provides a method for detecting the uniformity during the shaking bag packaging process of granular products. After the granular products are evenly mixed in the mixing stage, they enter the shaking bag packaging process. The method includes:

[0005] Obtaining the full-process spectrum of the target granular product during the mixing stage, and determining the mixing end point of the target granular product according to the full-process spectrum;

[0006] Collecting multiple spectral samples according to the mixing end point, and establishing a principal component analysis model according to the multiple spectral samples;

[0007] Determining the warning line and control line of the Hotelling statistic using the principal component analysis model and the multiple spectral samples;

[0008] Obtaining the real-time spectrum of the target granular product during the shaking bag packaging process, and detecting the uniformity of the granular product during the shaking bag packaging process using the principal component analysis model, the warning line, and the control line of the Hotelling statistic.

[0009] The method for detecting the uniformity during the shaking bag packaging process of the granular product provided by the present invention constructs a principal component analysis model by using the spectra in the mixing stage, and then uses the principal component analysis model to detect the uniformity of the granular product during the shaking bag packaging process, ensuring that the uniformity of the material after shaking bag mixing before bagging is consistent with that after passing the mixing stage, ensuring the quality of the finished product and improving the factory efficiency.

[0010] In an alternative embodiment, determining the mixing end point according to the full-process spectra includes:

[0011] Calculating the moving block standard deviation between each spectrum by using a preset algorithm;

[0012] Determining the mixing end point that meets the preset requirements according to the moving block standard deviation between each spectrum.

[0013] The method for detecting the uniformity during the shaking bag packaging process of the granular product provided by the present invention uses a preset algorithm to process the full-process spectra in the mixing stage in real time, discovers the mixing end point in time, avoids overmixing or under-mixing, and through real-time data feedback, can dynamically adjust the mixing parameters to ensure that the mixing process is always in the best state, accurately judge the mixing end point, which is beneficial to detecting the mixing state of the mixed material in the shaking bag packaging stage, ensuring the consistency of the material uniformity between the shaking bag packaging stage and the mixing stage, and improving the production efficiency and product quality.

[0014] In an alternative embodiment, collecting a plurality of spectral samples according to the mixing end point and establishing a principal component analysis model according to the spectral samples includes:

[0015] Obtaining the full-process spectra of the mixing stage of a plurality of batches of target granular products;

[0016] Selecting the spectra with a moving block standard deviation not greater than the mixing end point from the full-process spectra of the mixing stage of each batch of target granular products as spectral samples according to the mixing end point;

[0017] Performing model training by using a plurality of spectral samples to obtain a principal component analysis model.

[0018] In an alternative embodiment, determining the warning line and control line of the Hotelling statistic by using the principal component analysis model and a plurality of spectral samples includes:

[0019] Calculating the sample mean and sample covariance of a plurality of spectral samples by using the principal component analysis model;

[0020] Determining the warning line and control line of the Hotelling statistic of the target granular product according to the sample mean and sample covariance in the principal component space.

[0021] The present invention provides a method for detecting uniformity during the bag shaking and packaging process of a granular product. The method uses the spectrum near the mixing endpoint as a spectral sample, utilizes the spectral sample to train a principal component analysis model, and determines the control line of the mixing endpoint. The data volume of the spectral sample after dimension reduction is significantly reduced, which can speed up model training and prediction, improve computing efficiency, comprehensively analyze multiple variables in the spectrum, reveal the key change rules in the mixing process, and improve the accuracy of the mixing endpoint judgment.

[0022] In an optional embodiment, the uniformity of the granular product during the shaking bag packaging process is detected using a principal component analysis model and warning and control lines of Hotelling statistics, including:

[0023] The real-time spectrum is reduced in dimension using the principal component analysis model to obtain a reduced-dimensional matrix of the real-time spectrum;

[0024] The real-time Hotelling statistic is calculated based on the dimension reduction matrix of the real-time spectrum, and the uniformity of the granular product in the shaking bag packaging process is detected based on the real-time Hotelling statistic and the warning line and control line of the Hotelling statistic.

[0025] In an optional embodiment, the uniformity of the granular product during the bag shaking and packaging process is detected based on the real-time Hotelling statistic and the warning line and control line of the Hotelling statistic, including:

[0026] If the real-time Hotelling statistic is greater than the warning line of the Hotelling statistic and not greater than the control line of the Hotelling statistic, an early warning is issued;

[0027] If the real-time Hotelling statistic is greater than the control line of the Hotelling statistic, it means that the uniformity of the granular product in the bag shaking and packaging process is unqualified, and the bag shaking and packaging process is stopped.

[0028] The method for detecting uniformity of granular products during the bag shaking and packaging process provided by the present invention utilizes Hotelling statistics in combination with a principal component analysis model to more sensitively detect slight changes in uniformity, quickly identify abnormal batches or process fluctuations, standardize the uniformity detection process, reduce human errors, and promptly detect uniformity anomalies. Through real-time monitoring and early warning, the uniformity of each batch of products can be ensured, the sensitivity and accuracy of uniformity detection can be improved, and the quality of granular products can be improved.

[0029] In a second aspect, the present invention provides a device for detecting uniformity during the bag shaking and packaging process of a granular product. The granular product enters the bag shaking and packaging process after being uniformly mixed in the mixing stage. The device comprises:

[0030] A mixing endpoint determination module is used to obtain the full-process spectrum of the target particle product during the mixing stage and determine the mixing endpoint of the target particle product based on the full-process spectrum;

[0031] An analysis model establishment module, configured to collect multiple spectral samples according to a mixed endpoint and establish a principal component analysis model based on the multiple spectral samples;

[0032] A warning line and control line determination module, configured to determine a warning line and a control line of a Hotelling statistic by using the principal component analysis model and the multiple spectral samples;

[0033] An actual detection module, configured to obtain real-time spectra during the shaking bag and filling process of a target granular product, and perform uniformity detection on the granular product during the shaking bag and filling process by using the principal component analysis model, the warning line, and the control line of the Hotelling statistic.

[0034] In a third aspect, the present invention provides a computer device, including: a memory and a processor, which are communicatively connected to each other. The memory stores computer instructions, and the processor executes the computer instructions to execute the method according to the first aspect or any corresponding embodiment thereof.

[0035] In a fourth aspect, the present invention provides a computer-readable storage medium, on which computer instructions are stored. The computer instructions are used to cause a computer to execute the method according to the first aspect or any corresponding embodiment thereof.

[0036] In a fifth aspect, the present invention provides a computer program product, including computer instructions, which are used to cause a computer to execute the method according to the first aspect or any corresponding embodiment thereof. Description of the Drawings

[0037] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for use in the description of the specific embodiments or the prior art. Obviously, the following drawings are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0038] Figure 1 It is a schematic flowchart of a method for detecting the uniformity during the shaking bag and filling process of a granular product according to an embodiment of the present invention;

[0039] Figure 2 It is a schematic diagram of the detection position of a near-infrared spectrometer in a method for detecting the uniformity during the shaking bag and filling process of a granular product according to an embodiment of the present invention;

[0040] Figure 3 It is a schematic flowchart of another method for detecting the uniformity during the shaking bag and filling process of a granular product according to an embodiment of the present invention;

[0041] Figure 4It is a schematic diagram of the MBSD graph at the mixing end point in the method for detecting the uniformity during the shaking bag packaging process of granular products according to an embodiment of the present invention;

[0042] Figure 5 It is a schematic diagram of the Hotelling statistic, warning line, and control line during the mixing stage in a specific embodiment of the method for detecting the uniformity during the shaking bag packaging process of granular products according to an embodiment of the present invention;

[0043] Figure 6 It is a schematic diagram of the Hotelling statistic, warning line, and control line during the shaking bag stage in a specific embodiment of the method for detecting the uniformity during the shaking bag packaging process of granular products according to an embodiment of the present invention;

[0044] Figure 7 It is a structural block diagram of the device for detecting the uniformity during the shaking bag packaging process of granular products according to an embodiment of the present invention;

[0045] Figure 8 It is a schematic diagram of the hardware structure of the computer device according to an embodiment of the present invention. Detailed implementation manners

[0046] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0047] During the production process of granular products, the mixing of multiple materials is usually involved. Taking calcium carbonate D3 granules as an example, during the production process, after the intermediate and auxiliary materials are mixed with each other, they are packaged by the form of shaking bags to obtain the finished products. However, in the shaking bag packaging stage, the products filled in the bags may be uneven due to the inconsistent weights of the granules, resulting in unqualified finished products.

[0048] Based on the above problems, the embodiments of the present invention provide a method for detecting the uniformity during the shaking bag packaging process of granular products, which detects and warns the uniformity in the shaking bag packaging stage by analyzing the spectra in the mixing stage and the shaking bag packaging stage, so as to achieve the effect of ensuring the same uniformity after the shaking bag packaging stage and the mixing stage are qualified.

[0049] According to an embodiment of the present invention, an embodiment of a method for detecting the uniformity during the shaking bag packaging process of granular products is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than here.

[0050] In this embodiment, a method for detecting the uniformity during the shaking bag packaging process of granular products is provided, which can be used in the above computer system. Figure 1 It is a flowchart of the method for detecting the uniformity during the shaking bag packaging process of granular products according to an embodiment of the present invention. After the granular products are uniformly mixed in the mixing stage, they enter the shaking bag packaging process. As Figure 1 shown, this process includes the following steps:

[0051] Step S101, obtain the full-process spectrum of the target granular product in the mixing stage, and determine the mixing end point of the target granular product according to the full-process spectrum.

[0052] Specifically, as Figure 2 shown, it is a schematic diagram of the mixing stage and the shaking bag packaging stage of calcium carbonate D3 particles. When the intermediate and excipients are mixed in the mixing stage, the full-process spectrum of the materials during the mixing process is obtained in real time by using a near-infrared spectrometer. Initial stage of particle mixing: The particles are not fully mixed, and the spectral differences are large; Middle stage of mixing: The particles are gradually evenly distributed, and the spectral differences decrease; Late stage of mixing: The mixture tends to be uniform, and the spectral differences are stable at a low level. Therefore, the mixing end point of the target granular product can be determined by analyzing the differences between the spectra at adjacent sampling times. When the differences between the spectra at adjacent sampling times converge or are less than the preset threshold line, the mixing state of the target granular product is used as the mixing end point.

[0053] Step S102, collect a plurality of spectral samples according to the mixing end point, and establish a principal component analysis model based on the plurality of spectral samples.

[0054] Specifically, a plurality of spectra are collected near the sampling point of the mixing end point as spectral samples, and a principal component analysis model is established by using the spectral samples to analyze the characteristics of the mixed materials in the mixing end point state.

[0055] Step S103, determine the warning line and control line of the Hotelling statistic by using the principal component analysis model and the plurality of spectral samples.

[0056] Specifically, the principal component analysis model is used to perform dimensionality reduction processing on the plurality of spectral samples, and the value of the Hotelling statistic (Hotelling’s T 2 ) is calculated by using the dimensionality reduction data, so as to obtain the warning line and control line of the Hotelling statistic corresponding to the mixing end point in the mixing stage. The function of the warning line is that after calculating the actual Hotelling statistic according to the actually collected spectrum, if the actual Hotelling statistic exceeds the warning line, it indicates that the mixing uniformity of the granular product is not good, and a warning is issued to prompt the staff to take measures as soon as possible; the function of the control line is that if the actual Hotelling statistic exceeds the control line, it indicates that the mixing uniformity of the granular product is unqualified, and an unqualified prompt needs to be issued and the shaking bag packaging process is stopped.

[0057] Step S104: Obtain the real-time spectrum during the shaking and bagging process of the target granular product, and use the principal component analysis model, the warning line, and the control line of the Hotelling statistic to detect the uniformity of the granular product during the shaking and bagging process.

[0058] Specifically, the determined mixing end state in the mixing stage is used as the evaluation criterion for uniformity. The shaking and bagging process includes two steps: shaking the bag and filling the bag. During the process from shaking the bag to filling the bag, a real-time spectrum of the mixed material is obtained using another near-infrared spectrometer, and the real-time Hotelling statistic corresponding to the real-time spectrum is determined using the established principal component analysis model. The relationship between the real-time Hotelling statistic and the warning line and control line of the Hotelling statistic is compared, and whether the uniformity of the granular product during the shaking and bagging process meets the preset uniformity requirement is determined according to the comparison result.

[0059] The method for detecting the uniformity during the shaking and bagging process of the granular product provided in this embodiment constructs a principal component analysis model by using the spectrum in the mixing stage, and then uses the principal component analysis model to detect the uniformity of the granular product during the shaking and bagging process, ensuring that the uniformity of the material after shaking and mixing before bagging is consistent with that after passing the mixing stage, ensuring the quality of the finished product, and improving the factory efficiency.

[0060] In this embodiment, a method for detecting the uniformity during the shaking and bagging process of a granular product is provided, which can be used in the above computer system. Figure 3 is a flowchart of the method for detecting the uniformity during the shaking and bagging process of the granular product according to the embodiment of the present invention, as Figure 3 shown, and this process includes the following steps:

[0061] Step S201: Obtain the full-process spectrum of the target granular product in the mixing stage, and determine the mixing end point of the target granular product according to the full-process spectrum.

[0062] Specifically, the above step S201 includes:

[0063] Step S2011: Calculate the moving block standard deviation between each spectrum using a preset algorithm.

[0064] Specifically, the moving block standard deviation method (MBSD) is used to calculate the moving block standard deviation between the spectra at adjacent sampling times. The spectrum is represented by the mathematical formula as the near-infrared spectral data matrix X, with a size of m×p, where m is the number of samples and p is the number of wavelength points. The preset fixed moving block size in the moving block standard deviation method is set in advance as n. The specific calculation process for each spectral sample includes:

[0065] Calculate the moving block standard deviation for each wavelength point: For each wavelength point j (j = 1, 2,...., p), calculate the sum of squares, mean, variance, and standard deviation of the fixed moving block according to the following formula:

[0066] Sum of squares of the block S i,j :

[0067]

[0068] Block mean

[0069]

[0070] Variance of the block

[0071]

[0072] Standard deviation of the block σ i,j :

[0073]

[0074] After the calculation of the spectral samples covered by the moving block is completed, move the moving block backward to the adjacent spectral sample, and repeat the above calculation until all samples are covered. For each wavelength point j, obtain its moving block standard deviation sequence {σ 1,j , σ 2,j ,..., σ m-n+1，j}, according to the spectral standard deviation values, draw the MBSD mixed trend change line, determine the threshold line in combination with the mixed production requirements. In a certain calcium carbonate case, select the moving window block as 20 and draw the schematic diagram of the MBSD of the mixed end point of a certain batch as shown in Figure 4 Figure.

[0075] Step S2012, determine the mixing end point that meets the preset requirements according to the moving block standard deviation between each spectrum.

[0076] Specifically, sample at the near-infrared spectrometer, use the standard reference method corresponding to the sampled mixing method to measure the content of the target component in the spectral sample, and calculate the relative standard deviation (RSD). If it is ensured that RSD ≤ 2.0% within 1 minute, it is considered that the mixing end point is reached.

[0077] The method for detecting the uniformity during the vibrating bag packaging process of the granular product provided in this embodiment uses a preset algorithm to process the full-process spectrum in the mixing stage in real time, discovers the mixing end point in a timely manner, avoids overmixing or insufficient mixing, and through real-time data feedback, can dynamically adjust the mixing parameters to ensure that the mixing process is always in the best state, accurately judge the mixing end point, which is beneficial to detecting the mixing state of the mixed materials in the vibrating bag packaging stage, ensuring the consistency of the material uniformity between the vibrating bag packaging stage and the mixing stage, and improving production efficiency and product quality.

[0078] Step S202: Collect a plurality of spectral samples according to the mixing end point, and establish a principal component analysis model based on the plurality of spectral samples.

[0079] Specifically, the above step S202 includes:

[0080] Step S2021: Obtain the full-process spectrum of the mixing stage of a plurality of batches of target granular products.

[0081] Specifically, use a near-infrared spectrometer in the mixing stage to obtain the full-process spectrum of the mixing stage of a plurality of batches (generally more than 5 batches) of target granular products, and use each full-process spectrum as the original data.

[0082] Step S2022: Select the spectra with a moving block standard deviation not greater than the mixing end point from the full-process spectra of the mixing stage of each batch of target granular products as spectral samples according to the mixing end point.

[0083] Specifically, calculate the moving block standard deviation of the full-process spectra of each batch, determine the mixing end point of the corresponding batch according to the standard deviation, and based on the moving block standard deviation of each spectrum, select the spectra with a moving block standard deviation not greater than the mixing end point as spectral samples, that is, select the spectra near the mixing end point as spectral samples.

[0084] Step S2023: Use a plurality of spectral samples for model training to obtain a principal component analysis model.

[0085] Specifically, the specific process of using a plurality of spectral samples for model training includes the following steps:

[0086] Step a1: Standardize the plurality of spectral samples, and the standardization formula is:

[0087]

[0088] where X represents the original spectral data matrix, its size is n×p, n represents the number of samples, and p represents the number of spectral features (number of wavelengths); X stdDenote the standardized spectral data matrix, with size \(n\times p\); \(\mu\) represents the mean vector, with size \(1\times p\), denoting the mean of each spectral feature; \(\sigma\) represents the standard deviation vector, with size \(1\times p\), denoting the standard deviation of each spectral feature; \(\epsilon\) represents a very small regularization constant for numerical stability; \(I\) represents the identity matrix, with size \(n\times p\), for regularization.

[0089] Step a2, calculate the covariance matrix based on the standardized spectral data matrix:

[0090] X std = USV T

[0091]

[0092] where, \(\sum\) represents the covariance matrix, with size \(p\times p\), denoting the linear relationship between spectral features; \(U\) represents the left singular vector matrix, with size \(n\times n\), denoting the basis vectors of the sample space; \(S\) represents the singular value diagonal matrix, with size \(n\times p\), and the elements on the diagonal are singular values; \(V\) represents the right singular vector matrix, with size \(p\times p\), denoting the basis vectors of the feature space; \(V^T\) represents the transpose of the right singular vector matrix, with size \(p\times p\); \(n - 1\) represents the degrees of freedom correction term for unbiased estimation.

[0093] Step a3, perform eigenvalue decomposition on the covariance matrix:

[0094] \(\sum = Q\Lambda Q^T\) (7) T (7)

[0095] Q orth = Gram - Schmidt(Q) (8)

[0096] where, \(Q\) represents the eigenvector matrix, with size \(p\times p\), and each column is an eigenvector, denoting the direction of the principal component; \(\Lambda\) represents the eigenvalue diagonal matrix, with size \(p\times p\), and the elements on the diagonal are the eigenvalues \(\lambda_1,\lambda_2,\cdots,\lambda_p\), denoting the differences of the principal components; \(Q^T\) represents the transpose of the eigenvector matrix, with size \(p\times p\); \(\hat{Q}\) T represents the transpose of the eigenvector matrix, with size \(p\times p\); \(\hat{Q}\) orth represents the eigenvector matrix after Gram - Schmidt orthogonalization, with size \(p\times p\).

[0097] Step a4, select the principal components from the eigenvector matrix after Gram - Schmidt orthogonalization:

[0098] \(\hat{Q}\) k = \(\hat{Q}\) orth [:, 1:k]\(\cdot W\) (9)

[0099] where, \(\hat{Q}\) kDenote the projection matrix, with size p×k, representing the directions of the first k principal components; Q orth [:, 1:k] represents the first k columns of the orthonormalized eigenvector matrix, with size p×k; W represents the diagonal weight matrix, with size k×k, and the elements on the diagonal are weights w1, w2,...., w i . The weights are usually set as: That is, the normalized weights of the eigenvalues, and k represents the number of dimensions after dimensionality reduction (k < p).

[0100] Step a5, project the principal components:

[0101] Y = X std Q k + B (10)

[0102] where Y represents the data matrix after dimensionality reduction, with size n×k, representing the coordinates of the samples in the principal component space; X std represents the standardized spectral data matrix, with size n×p; Q k represents the projection matrix, with size p×k; B represents the bias matrix, with size n×k, usually set as B = ∈I, where ∈ is a very small constant.

[0103] Step a6, inverse transformation calculation:

[0104]

[0105] where X reconstructed represents the reconstructed spectral data matrix, with size n×p; represents the transpose of the projection matrix, with size k×p; ∈I represents a very small regularization term, with size n×p.

[0106] After dimensionality reduction processing of the spectral samples, the principal components related to the material mixing uniformity are obtained. Using the projection matrix of the principal components to train the general principal component analysis model, a principal component analysis model suitable for the target particulate product is obtained, reducing the data volume and improving the efficiency of uniformity analysis on the premise of ensuring the accuracy of uniformity analysis.

[0107] During actual prediction, after dimensionality reduction of the spectral data, it is substituted into formula (11) for prediction.

[0108] Step S203, determine the warning line and control line of the Hotelling statistic using the principal component analysis model and multiple spectral samples.

[0109] Specifically, the above step S203 includes:

[0110] Step S2031, calculate the sample mean and sample covariance of multiple spectral samples using the principal component analysis model.

[0111] Specifically, calculate the sample covariance of multiple spectral samples according to formula (6) in the principal component analysis model, and calculate the sample mean The formula is:

[0112]

[0113] where y: the coordinate of a single sample in the principal component space (k×1);

[0114] Calculate the covariance matrix in the principal component space:

[0115]

[0116] Step S2032, determine the warning line and control line of the Hotelling statistic of the target granular product according to the sample mean and sample covariance in the principal component space.

[0117] Specifically, substitute the sample mean and sample covariance into the calculation formula of the preset Hotelling statistic:

[0118]

[0119] where represents the mean vector of the training set in the principal component space, and Σ represents the sample covariance.

[0120] Use Hotelling's T-squared to calculate the F statistic:

[0121]

[0122] where n represents the number of samples participating in the modeling, that is, the product of the number of collection batches and the number of intercepted time points in each batch; F α (k, n-k) represents the critical value of the F distribution corresponding to the significance level α.

[0123] P{F(n1, n2)>F α (n1, n2)} = α (14)

[0124] where 1-α represents the confidence level. When α is 0.05, it represents the warning line, which means a 95% confidence interval; when α is 0.01, it represents the control line, which means a 99% confidence interval.

[0125] The method for detecting the uniformity during the shaking bag packaging process of the granular product provided by the present embodiment is based on the spectrum near the mixing end point as the spectral sample, uses the spectral sample to train the principal component analysis model and determines the control line of the mixing end point. The data volume after the spectral sample is dimensionally reduced is significantly reduced, which can accelerate the training and prediction speed of the model, improve the calculation efficiency, comprehensively analyze multiple variables in the spectrum, reveal the key change rules during the mixing process, and improve the accuracy of judging the mixing end point.

[0126] In step S204, obtain the real-time spectrum during the vibrating bag filling process of the target granular product, and use the principal component analysis model, the warning line and the control line of the Hotelling statistic to detect the uniformity of the granular product during the vibrating bag filling process.

[0127] Specifically, the above step S204 includes:

[0128] In step S2041, use the principal component analysis model to perform dimensionality reduction processing on the real-time spectrum to obtain the dimensionality reduction matrix of the real-time spectrum.

[0129] Specifically, use the near-infrared spectrometer in the vibrating bag filling stage to obtain the real-time spectrum during the vibrating bag filling process, and use the principal component analysis model to perform dimensionality reduction processing on the real-time spectrum to obtain the dimensionality reduction matrix of the real-time spectrum. For the specific process, please refer to steps a1-a5 and will not be elaborated here.

[0130] In step S2042, calculate the real-time Hotelling statistic according to the dimensionality reduction matrix of the real-time spectrum, and detect the uniformity of the granular product during the vibrating bag filling process according to the real-time Hotelling statistic and the warning line and the control line of the Hotelling statistic.

[0131] Specifically, calculate the real-time Hotelling statistic according to the dimensionality reduction matrix of the real-time spectrum, and detect the uniformity of the granular product during the vibrating bag filling process according to the real-time Hotelling statistic and the warning line and the control line of the Hotelling statistic.

[0132] In some optional embodiments, detecting the uniformity of the granular product during the vibrating bag filling process according to the real-time Hotelling statistic and the warning line and the control line of the Hotelling statistic includes:

[0133] In step b1, if the real-time Hotelling statistic is greater than the warning line of the Hotelling statistic and not greater than the control line of the Hotelling statistic, issue a warning.

[0134] Specifically, determine the warning line and the control line of the Hotelling statistic by using the spectral samples near the mixing end point in the mixing stage. Add a near-infrared spectrometer during the process from vibrating bag to bag filling to detect the uniformity of the material before bag filling, so as to know the uniformity of the material when bag filling.

[0135] If the real-time Hotelling statistic is greater than the warning line of the Hotelling statistic and not greater than the control line of the Hotelling statistic, it indicates that there is a gap between the uniformity of the granular product before bag filling and the bag filling requirements, but the gap will not cause the product after bag filling to be unqualified. Then issue a warning to prompt the staff to take measures to improve the uniformity during the vibrating bag filling process.

[0136] Step b2, if the real-time Hotelling statistic is greater than the control line of the Hotelling statistic, it indicates that the uniformity of the granular product in the shaking bag filling process is unqualified, and the shaking bag filling process is stopped.

[0137] Specifically, if the real-time Hotelling statistic is greater than the control line of the Hotelling statistic, it indicates that the uniformity of the granular product in the shaking bag filling process is unqualified, then a prompt can be given in time and the shaking bag filling process can be stopped to avoid subsequent problems caused by uneven finished products being bagged and shipped out.

[0138] The uniformity detection method for the shaking bag filling process of the granular product provided in this embodiment can more sensitively detect small changes in uniformity by using the Hotelling statistic combined with the principal component analysis model, quickly identify abnormal batches or process fluctuations, standardize the uniformity detection process, reduce human errors, and timely detect uniformity abnormalities. Through real-time monitoring and early warning, it can ensure the consistency of the uniformity of each batch of products, improve the sensitivity and accuracy of uniformity detection, and improve the quality of granular products.

[0139] In a specific embodiment, a near-infrared spectrometer is installed on the mixing tank to collect spectra during the mixing stage of calcium carbonate D3 granule production. The collection time interval is 3 s. According to the previous mixing production research, the production MBSD threshold line is determined to be 0.02, and 1 min after reaching the threshold line is the mixing end point.

[0140] Import the full-process spectral data of the mixing stage into the Unscrambler software, automatically calculate the MBSD value of the mixing stage, collect the spectra for 1 min after reaching the threshold line, and collect the production spectra of 7 batches in total. Import the production spectra of the 7 batches into the Unscrambler software respectively, and use moving average - second derivative - SNV preprocessing to reduce the influence brought by the spectral background, improve the difference between spectra, and improve the spectral recognition. Perform PCA dimensionality reduction analysis on the preprocessed spectral data, establish a PCA model, and calculate the Hotelling’s T 2 value of the spectra of 7 batches to obtain the warning line and control line of the standard production Hotelling’s T 2 As shown in Figure 5 the schematic diagram of the warning line and control line in the Hotelling statistical data (Hotelling’s T 2 Statistic) obtained according to the Hotelling T 2 distribution in the standard production of the mixing stage.

[0141] The collection parameters of the near-infrared spectrometer installed in the shaking bag stage are kept consistent with those in the mixing stage, collect the spectra of the bagged samples, and use the established PCA model for prediction to obtain the Hotelling’s T 2 value of the spectra in the shaking bag stage, based on the standard production Hotelling’s T2 The warning line and the control line are used to examine whether the uniformity during the vibrating bag filling stage meets the production standards. For example, Figure 6 As shown in 2 the schematic diagram of the change of Hotelling’s T 2 in the Hotelling’s T

[0142] In this embodiment, a device for detecting the uniformity during the vibrating bag filling process of granular products is also provided. This device is used to implement the above-mentioned embodiments and preferred implementation manners, and those that have been described will not be repeated. As used below, the term "module" can be a combination of software and / or hardware that can achieve a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, implementation in hardware, or a combination of software and hardware is also possible and contemplated.

[0143] This embodiment provides a device for detecting the uniformity during the vibrating bag filling process of granular products. As Figure 7 shown, it includes:

[0144] A mixing end determination module 601, configured to obtain the full-process spectrum of the target granular product during the mixing stage, and determine the mixing end of the target granular product according to the full-process spectrum.

[0145] An analysis model establishment module 602, configured to collect a plurality of spectral samples according to the mixing end, and establish a principal component analysis model according to the plurality of spectral samples.

[0146] A warning line and control line determination module 603, configured to determine the warning line and control line of the Hotelling statistic by using the principal component analysis model and the plurality of spectral samples.

[0147] An actual detection module 604, configured to obtain the real-time spectrum during the vibrating bag filling process of the target granular product, and perform uniformity detection on the granular product during the vibrating bag filling process by using the principal component analysis model, the warning line, and the control line of the Hotelling statistic.

[0148] In some alternative implementation manners, the mixing end determination module 601 includes:

[0149] A standard deviation calculation unit, configured to calculate the moving block standard deviation between each spectrum by using a preset algorithm.

[0150] A mixing end determination unit, configured to determine the mixing end that meets the preset requirements according to the moving block standard deviation between each spectrum.

[0151] In some alternative implementation manners, the analysis model establishment module 602 includes:

[0152] A spectrum acquisition unit for acquiring the full-process spectra of multiple batches of target particle products during the mixing stage.

[0153] A spectrum sample calculation unit for selecting, based on the mixing endpoint, spectra with a moving block standard deviation not greater than the mixing endpoint from the full-process spectra of each batch of target particle products during the mixing stage as spectrum samples.

[0154] A model training unit for training a principal component analysis model using multiple spectrum samples.

[0155] In some alternative embodiments, the warning line and control line determination module 603 includes:

[0156] A spectrum sample dimensionality reduction unit for calculating the sample mean and sample covariance of multiple spectrum samples using the principal component analysis model.

[0157] A warning line and control line determination unit for determining the warning line and control line of the Hotelling statistic of the target particle product based on the sample mean and sample covariance.

[0158] In some alternative embodiments, the actual detection module 604 includes:

[0159] A spectrum dimensionality reduction unit for performing dimensionality reduction processing on the real-time spectrum using the principal component analysis model to obtain the dimensionality reduction matrix of the real-time spectrum.

[0160] A uniformity detection unit for calculating the real-time Hotelling statistic based on the dimensionality reduction matrix of the real-time spectrum, and performing uniformity detection on the particle products during the shaking bag filling process based on the real-time Hotelling statistic and the warning line and control line of the Hotelling statistic.

[0161] In some alternative embodiments, the uniformity detection unit includes:

[0162] A detection warning subunit for issuing a warning if the real-time Hotelling statistic is greater than the warning line of the Hotelling statistic and not greater than the control line of the Hotelling statistic.

[0163] A detection stop subunit for indicating that the uniformity of the particle products during the shaking bag filling process is unqualified and stopping the shaking bag filling process if the real-time Hotelling statistic is greater than the control line of the Hotelling statistic.

[0164] The further function descriptions of the above-mentioned various modules and units are the same as those in the corresponding embodiments above, and will not be elaborated here.

[0165] In this embodiment, the uniformity detection device during the shaking bag packaging process of granular products is presented in the form of functional units. Here, the unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and a memory that execute one or more software or fixed programs, and / or other devices that can provide the above functions.

[0166] An embodiment of the present invention further provides a computer device having the above Figure 7 uniformity detection device during the shaking bag packaging process of granular products as shown.

[0167] Please refer to Figure 8 , Figure 8 which is a schematic structural diagram of a computer device provided by an alternative embodiment of the present invention. As shown in Figure 8 , the computer device includes: one or more processors 10, a memory 20, and interfaces for connecting various components, including a high-speed interface and a low-speed interface. Each component communicates with each other using different buses and can be installed on a common motherboard or installed in other ways as needed. The processor can process instructions executed within the computer device, including instructions stored in the memory or on the memory to display graphical information of the GUI on an external input / output device (such as a display device coupled to the interface). In some alternative embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Similarly, multiple computer devices can be connected, and each device provides some necessary operations (for example, as a server array, a set of blade servers, or a multi-processor system). Figure 8 In

[0168] FIG.

[0169] The memory 20 stores instructions executable by at least one processor 10, so that the at least one processor 10 executes the method shown in the above embodiment.

[0170] The memory 20 may include a program storage area and a data storage area. Among them, the program storage area can store an operating system and application programs required for at least one function; the data storage area can store data created according to the use of the computer device and the like. In addition, the memory 20 may include high-speed random access memory, and may also include non-transitory memory, such as at least one magnetic disk storage device, a flash memory device, or other non-transitory solid-state storage devices. In some alternative embodiments, the memory 20 may optionally include a memory remotely provided relative to the processor 10, and these remote memories can be connected to the computer device through a network. Examples of the above-mentioned network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0171] The memory 20 may include volatile memory, for example, random access memory; the memory may also include non-volatile memory, for example, flash memory, a hard disk, or a solid-state drive; the memory 20 may further include a combination of the above types of memory.

[0172] The computer device further includes a communication interface 30 for the computer device to communicate with other devices or a communication network.

[0173] The embodiments of the present invention further provide a computer-readable storage medium. The methods according to the embodiments of the present invention can be implemented in hardware, firmware, or be implemented as computer code that can be recorded on a storage medium, or be implemented as computer code originally stored in a remote storage medium or a non-transitory machine-readable storage medium and downloaded through a network and to be stored in a local storage medium, so that the methods described herein can be stored in such software processes on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only memory, a random access memory, a flash memory, a hard disk, or a solid-state drive, etc.; further, the storage medium may also include a combination of the above types of memory. It can be understood that a computer, a processor, a microprocessor controller, or programmable hardware includes a storage component that can store or receive software or computer code, and when the software or computer code is accessed and executed by the computer, the processor, or the hardware, the methods shown in the above embodiments are implemented.

[0174] A part of the present invention can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can call or provide the methods and / or technical solutions according to the present invention through the operations of the computer. Those skilled in the art should understand that the forms of existence of computer program instructions in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executes the instructions, or the computer compiles the instructions and then executes the corresponding compiled program, or the computer reads and executes the instructions, or the computer reads and installs the instructions and then executes the corresponding installed program. Herein, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to the computer.

[0175] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A method for detecting the uniformity during the shaking bag packaging process of granular products. After the granular products are evenly mixed in the mixing stage, they enter the shaking bag packaging process. It is characterized in that, The method includes: Obtaining the full-process spectrum of the mixing stage of the target granular product, and determining the mixing end point of the target granular product according to the full-process spectrum; Collecting a plurality of spectral samples according to the mixing end point, and establishing a principal component analysis model according to the plurality of spectral samples; Determining the warning line and control line of the Hotelling statistic by using the principal component analysis model and the plurality of spectral samples; Obtaining the real-time spectrum during the shaking bag and filling process of the target granular product, and detecting the uniformity of the granular product during the shaking bag and filling process by using the principal component analysis model, the warning line and the control line of the Hotelling statistic.

2. The method according to claim 1, wherein Determining the mixing end point according to the full-process spectrum includes: Calculating the moving block standard deviation between each spectrum by using a preset algorithm; Determining the mixing end point that meets the preset requirements according to the moving block standard deviation between each spectrum.

3. The method according to claim 2, wherein Collecting a plurality of spectral samples according to the mixing end point, and establishing a principal component analysis model according to the spectral samples, includes: Obtaining the full-process spectra of the mixing stages of a plurality of batches of target granular products; Selecting, as spectral samples, the spectra with a moving block standard deviation not greater than the mixing end point from the full-process spectra of the mixing stages of each batch of target granular products according to the mixing end point; Performing model training by using the plurality of spectral samples to obtain a principal component analysis model.

4. The method according to claim 3, wherein The determining the warning line and control line of the Hotelling statistic by using the principal component analysis model and the plurality of spectral samples includes: Calculating the sample mean and sample covariance of the plurality of spectral samples by using the principal component analysis model; Determining the warning line and control line of the Hotelling statistic of the target granular product according to the sample mean and sample covariance in the principal component space.

5. The method according to claim 1, characterized in that, Detecting the uniformity of the granular product during the shaking bag and filling process by using the principal component analysis model, the warning line and the control line of the Hotelling statistic includes: Performing dimensionality reduction processing on the real-time spectrum by using the principal component analysis model to obtain the dimensionality reduction matrix of the real-time spectrum; Calculating the real-time Hotelling statistic according to the dimensionality reduction matrix of the real-time spectrum, and detecting the uniformity of the granular product during the shaking bag and filling process according to the real-time Hotelling statistic and the warning line and control line of the Hotelling statistic.

6. The method according to claim 5, wherein Detecting the uniformity of the granular product during the shaking bag and filling process according to the real-time Hotelling statistic and the warning line and control line of the Hotelling statistic includes: If the real-time Hotelling statistic is greater than the warning line of the Hotelling statistic and not greater than the control line of the Hotelling statistic, giving an early warning; If the real-time Hotelling statistic is greater than the control line of the Hotelling statistic, it indicates that the uniformity of the granular product during the shaking bag and filling process is unqualified, and the shaking bag and filling process is stopped.

7. A uniformity detection device during the shaking bag packaging process of granular products. After the granular products are evenly mixed in the mixing stage, they enter the shaking bag packaging process. It is characterized in that, The device includes: A mixing end point determination module, configured to obtain the full-process spectrum of the mixing stage of the target granular product, and determine the mixing end point of the target granular product according to the full-process spectrum; An analysis model establishment module, configured to collect a plurality of spectral samples according to the mixing end point, and establish a principal component analysis model according to the plurality of spectral samples; A warning line and control line determination module, configured to determine the warning line and control line of the Hotelling statistic by using the principal component analysis model and multiple spectral samples; An actual detection module, configured to obtain the real-time spectrum during the shaking bag and filling process of the target granular product, and perform uniformity detection on the granular product during the shaking bag and filling process by using the principal component analysis model, the warning line and the control line of the Hotelling statistic.

8. A computer device, characterized in that, Comprising: A memory and a processor, which are communicatively connected to each other. The memory stores computer instructions, and the processor executes the computer instructions to execute the method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, Computer instructions are stored on the computer-readable storage medium, and the computer instructions are used to cause a computer to execute the method according to any one of claims 1 to 6.

10. A computer program product, characterized in that, Comprising computer instructions, which are used to cause a computer to execute the method according to any one of claims 1 to 6.

Images