Monopotassium phosphate quality control method based on big data analysis

Through clustering and data correction technology, the long-term dependence problem of recurrent neural networks in the production of potassium dihydrogen phosphate is solved, the prediction efficiency and production control effect are improved, and the high-quality production of potassium dihydrogen phosphate is ensured.

CN119069016AActive Publication Date: 2024-12-03SHANDONG PROVINCE DINGXIN BIOLOGY TECH CO LTD

Patent Information

Application Number
CN202411177194.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-26
Publication Date
2024-12-03
Estimated Expiration
2044-08-26

AI Technical Summary

Technical Problem

The recurrent neural network has long-term dependence problems in the production process of potassium dihydrogen phosphate, which cannot effectively meet the real-time requirements, resulting in insufficient prediction efficiency and the inability to control the production quality of potassium dihydrogen phosphate in a timely manner.

Method used

By obtaining the reaction data of current and historical potassium dihydrogen phosphate, clustering, screening the approximate reaction clusters, supplementing and correcting the reaction data, combining representative quality inspection data and neural networks for quality prediction, and adjusting production parameters through decision trees, realizing Quality control.

Benefits of technology

It reduces the long-term dependence problem of neural networks, improves prediction efficiency and controllability of the production process, and ensures high-quality production of potassium dihydrogen phosphate.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119069016A_ABST
    Figure CN119069016A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of electrical digital data processing, in particular to a monopotassium phosphate quality control method based on big data analysis, which comprises the following steps: dividing historical monopotassium phosphate into a plurality of clusters, screening out an approximate reaction cluster according to the similar condition between the current monopotassium phosphate and the historical monopotassium phosphate in different clustering clusters, and supplementing and correcting the reaction data of the current monopotassium phosphate after the latest moment; and on the basis of the distribution condition of the quality inspection data in each cluster, representative quality inspection data of each cluster is obtained, and parameters in the current monopotassium phosphate production process are adjusted by combining the representative quality inspection data and utilizing a neural network and a decision tree, so that the quality control of the current monopotassium phosphate is realized. According to the method, the analysis and prediction efficiency of the neural network on the monopotassium phosphate in the real-time production process is improved, and the controllability and control efficiency of the monopotassium phosphate in the production process are improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the technical field of electrical digital data processing, and in particular to a potassium dihydrogen phosphate quality control method based on big data analysis. Background Art

[0002] Potassium dihydrogen phosphate is an inorganic phosphate with a wide range of applications. In order to meet actual needs, increasing the output of high-quality potassium dihydrogen phosphate is of great practical significance. However, the current production and preparation process of potassium dihydrogen phosphate is affected by a variety of factors, and there is a relationship of mutual influence between the various factors, which makes the physical and chemical reaction environment in the production process complex, resulting in fluctuations in the production quality of potassium dihydrogen phosphate.

[0003] Therefore, in order to ensure the production quality of potassium dihydrogen phosphate, the various data in the production process of potassium dihydrogen phosphate are currently analyzed and predicted by combining big data technology, so as to adjust the various parameters of potassium dihydrogen phosphate in the production process in real time, thereby ensuring the production quality. However, when using recurrent neural network (RNN) to predict the quality of the potassium dihydrogen phosphate production process, RNN has certain long-term dependence problems and cannot effectively meet the real-time requirements, resulting in insufficient prediction efficiency of RNN, and unable to effectively and timely control the production quality of potassium dihydrogen phosphate. Summary of the invention

[0004] The present invention provides a potassium dihydrogen phosphate quality control method based on big data analysis to solve the existing problems.

[0005] The potassium dihydrogen phosphate quality control method based on big data analysis of the present invention adopts the following technical scheme:

[0006] One embodiment of the present invention provides a potassium dihydrogen phosphate quality control method based on big data analysis, the method comprising the following steps:

[0007] Obtain some reaction data of current and historical potassium dihydrogen phosphate, as well as historical quality inspection data of potassium dihydrogen phosphate;

[0008] Based on the similarity of the quality inspection data, the corresponding historical potassium dihydrogen phosphate is divided into several clusters, and the approximate reaction clusters of the current potassium dihydrogen phosphate are screened out according to the similarity in the change trend between the reaction data of the current potassium dihydrogen phosphate and the historical potassium dihydrogen phosphate in different clusters;

[0009] The reaction data after the latest moment of the current potassium dihydrogen phosphate is supplemented by using the correlation between the current potassium dihydrogen phosphate and the historical potassium dihydrogen phosphate before the latest moment, so as to obtain the new reaction data of the current potassium dihydrogen phosphate, and the new reaction data is corrected by using the slope and value of each data point in the new reaction data and the correlation between the new reaction data and the reaction data of the historical potassium dihydrogen phosphate in the approximate reaction cluster, so as to obtain the final reaction data of the current potassium dihydrogen phosphate;

[0010] Based on the distribution of quality inspection data in each cluster, obtain representative quality inspection data of each cluster;

[0011] The quality of the current potassium dihydrogen phosphate is predicted by combining representative quality inspection data and using a neural network. The parameters in the current potassium dihydrogen phosphate production process are adjusted through a decision tree to achieve quality control of the current potassium dihydrogen phosphate.

[0012] Furthermore, the similarity in the change trend between the current potassium dihydrogen phosphate and the reaction data of the historical potassium dihydrogen phosphate in different clusters is used to screen out the approximate reaction cluster of the current potassium dihydrogen phosphate, including the specific method of:

[0013] Based on the change trends of the reaction data of the current potassium dihydrogen phosphate and the historical potassium dihydrogen phosphate in each cluster, the trend information sequence of the current potassium dihydrogen phosphate and the trend information sequence of each cluster are obtained respectively;

[0014] The similarity between the trend information sequence of the current potassium dihydrogen phosphate and the historical potassium dihydrogen phosphate in each cluster is analyzed to obtain the reaction similarity between the current potassium dihydrogen phosphate and each cluster, and the cluster corresponding to the maximum reaction similarity is taken as the approximate reaction cluster of the current potassium dihydrogen phosphate.

[0015] Furthermore, the specific method for obtaining the trend information sequence of the cluster is:

[0016] Perform STL time series decomposition on all historical potassium dihydrogen phosphate reaction data contained in any cluster to obtain the corresponding trend item sequence; construct a two-dimensional rectangular coordinate system, place the potassium dihydrogen phosphate reaction data trend item sequence corresponding to all quality inspection data of the same cluster into the same two-dimensional rectangular coordinate system, use the least squares method to perform curve fitting on all trend item sequences in the two-dimensional rectangular coordinate system, and obtain the fitting curve of each cluster, which is recorded as the reaction trend curve of the cluster;

[0017] Obtain several extreme points of any reaction trend curve, divide the reaction trend curve into several curve segments using the extreme points, obtain the time corresponding to the extreme points at both ends of the curve segment, and perform straight line fitting on the curve segments between any adjacent extreme values ​​to obtain the slope and intercept of the corresponding straight line, obtain an array formed by the corresponding time of the extreme points at both ends of all curve segments of any reaction trend curve, and an array formed by the slope and intercept of the corresponding fitting straight line of all curve segments, which are respectively recorded as the time array and the straight line parameter array of the curve segment, and obtain a sequence formed by all the time arrays and the straight line parameter arrays in the reaction trend curve according to the time sequence of each curve segment in the reaction trend curve, which is recorded as the trend information sequence of the corresponding cluster cluster.

[0018] Furthermore, the specific method for obtaining the trend information sequence of the current potassium dihydrogen phosphate is:

[0019] Perform STL time series decomposition on any reaction data of potassium dihydrogen phosphate in the current production process, obtain the corresponding trend item and perform curve fitting, obtain the corresponding reaction trend curve, obtain the extreme points of the reaction trend curve and divide it into several curve segments, obtain the sequence formed by the time array and the straight line parameter array of each curve segment, so as to obtain the corresponding trend information sequence of the current potassium dihydrogen phosphate.

[0020] Furthermore, the specific calculation method of the reaction similarity is:

[0021]

[0022] Among them, f a is the reaction similarity between the current potassium dihydrogen phosphate and the a-th cluster; x ji represents the i-th element of the j-th trend information sequence of the current potassium dihydrogen phosphate, x′ aji represents the ith element of the jth trend information sequence of the ath cluster, J represents the number of trend information sequences, d() represents obtaining the Euclidean distance, I represents the quantity parameter, I″ represents the number of elements in the trend information sequence corresponding to the current potassium dihydrogen phosphate, and I′ a It indicates the number of elements in the trend information sequence corresponding to the a-th cluster, and min() means obtaining the minimum value.

[0023] Furthermore, the specific method for obtaining the new reaction data of the current potassium dihydrogen phosphate is:

[0024] For any reaction data of current potassium dihydrogen phosphate, obtain the latest time of the reaction data;

[0025] Using the latest time, the corresponding reaction data of all historical potassium dihydrogen phosphate in the approximate reaction cluster are divided into two sections, one is the reaction data section before the latest time, and the other is the reaction data section after the latest time;

[0026] According to the similarity relationship between the reaction data of the current potassium dihydrogen phosphate and different reaction data segments, and in combination with the reaction data segments after the latest reaction of all historical potassium dihydrogen phosphates, the reaction data of the current potassium dihydrogen phosphate is completed to obtain new reaction data of the current potassium dihydrogen phosphate, wherein the new reaction data includes new temperature time series data, new pH time series data and new stirring speed time series data.

[0027] Furthermore, the specific method for obtaining the final reaction data of the current potassium dihydrogen phosphate is:

[0028] Obtaining a correction coefficient for each data point in the new reaction data according to the slope and size of each data point in any new reaction data and the correlation between the new reaction data and the reaction data of historical potassium dihydrogen phosphate in the approximate reaction cluster;

[0029] The new reaction data is corrected using the correction coefficient to obtain the final reaction data of the current potassium dihydrogen phosphate.

[0030] Furthermore, the specific calculation method of the correction coefficient of each data point in the new reaction data is:

[0031]

[0032] Among them, α1 represents the new temperature time series data, α2 represents the new pH time series data, α3 represents the new stirring speed time series data, γ(α u ) represents the correction coefficient of the u-th data point in the new reaction data α, represents the approximation factor of the new reaction data α, δ u () represents the numerical parameter of the uth data point of the new reaction data, δ1′ u represents the appropriate numerical parameter of the u-th data point in the new temperature time series data, || represents obtaining the absolute value, norm() represents the normalization function, exp() represents the exponential function with the natural constant as the base; the numerical parameter is the product of the slope of the data point in the new reaction data and the numerical value; the appropriate numerical parameter is the product of the slope of the data point in the new temperature time series data and Br, and Br is the appropriate temperature of potassium dihydrogen phosphate in the reactor during the reaction process.

[0033] Furthermore, the method of obtaining representative quality inspection data of each cluster based on the distribution of quality inspection data in each cluster includes the following specific methods:

[0034] For any cluster, the Euclidean distance between each quality inspection data in the cluster and the cluster center of the cluster is obtained, and all quality inspection data in the cluster are weightedly fused using the Euclidean distance to obtain representative quality inspection data of the cluster.

[0035] Furthermore, the new reaction data is corrected by using the correction coefficient to obtain the final reaction data of the current potassium dihydrogen phosphate, including the specific method of:

[0036] The calculation method for the value of the data point after the latest time in the final reaction data is:

[0037] L′(α u )=γ(α u )×L(α u )

[0038] Among them, L(α′ u ) represents the value of the data point corresponding to the uth time after the latest time in the final reaction data α′, γ(α u ) represents the correction coefficient of the data point corresponding to the uth time after the latest time in the new reaction data α, L(α u ) represents the value of the data point corresponding to the uth moment after the latest moment in the new reaction data α.

[0039] The beneficial effect of the technical solution of the present invention is as follows: the embodiment of the present invention takes into account that potassium dihydrogen phosphate is affected by various factors during the production process, so that there are certain differences in the quality inspection results of the potassium dihydrogen phosphate produced historically. Therefore, the embodiment of the present invention uses this difference to divide the historical potassium dihydrogen phosphate, and based on the similarity between the current potassium dihydrogen phosphate and the historical potassium dihydrogen phosphate in the reaction data, the reaction data of the current potassium dihydrogen phosphate is supplemented and corrected, and the commonality of the quality inspection data of the historical potassium dihydrogen phosphate in different clusters is combined to obtain representative quality inspection data to improve the accuracy of the neural network prediction results and reduce the amount of calculation in the neural network training process. In addition, by performing trend analysis on the reaction data as long time series data, the reaction data is converted into a trend information sequence under the premise of ensuring that the trend change information is not lost, which greatly reduces the influence of the long-term dependence problem of the neural network on the prediction results of the neural network, improves the efficiency of the neural network in analyzing and predicting potassium dihydrogen phosphate in the real-time production process, and improves the controllability and control efficiency of potassium dihydrogen phosphate in the production process. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0041] Figure 1 The present invention is a flowchart of the steps of the potassium dihydrogen phosphate quality control method based on big data analysis;

[0042] Figure 2 This is a schematic diagram of the quality control process of potassium dihydrogen phosphate. DETAILED DESCRIPTION

[0043] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the potassium dihydrogen phosphate quality control method based on big data analysis proposed by the present invention, its specific implementation method, structure, characteristics and effects, as follows in conjunction with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0044] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0045] The specific scheme of the potassium dihydrogen phosphate quality control method based on big data analysis provided by the present invention is described in detail below with reference to the accompanying drawings.

[0046] See also Figure 1 , which shows a flow chart of the steps of a potassium dihydrogen phosphate quality control method based on big data analysis provided by an embodiment of the present invention, the method comprising the following steps:

[0047] Step S001: Acquire some reaction data of current potassium dihydrogen phosphate and historical potassium dihydrogen phosphate, as well as quality inspection data of historical potassium dihydrogen phosphate.

[0048] It should be noted that the wet method of preparing potassium dihydrogen phosphate is usually based on the principle of acid-base neutralization. For example, in the currently commonly used neutralization method, potassium hydroxide and phosphoric acid are neutralized in a certain proportion to obtain potassium dihydrogen phosphate and water. After the reaction is completed, it needs to go through filtration, crystallization, centrifugal separation, drying and other processes. In each production process, there are corresponding influencing factors, including: ingredients: ensure the accuracy of the ingredient ratio and ingredient concentration; reaction and crystallization: control the reaction temperature and stirring speed; centrifugal separation: select appropriate operating parameters; drying: control the drying temperature and time.

[0049] Since the changes in various parameters inside the reactor during the wet preparation of potassium dihydrogen phosphate play an important role in the quality control of potassium dihydrogen phosphate, in order to better control the output quality of the reaction link in the production process of potassium dihydrogen phosphate, it is necessary to analyze the reaction data of the current production batch and the reaction data of the historical production batches, so as to control the production quality of potassium dihydrogen phosphate by adjusting the parameters in the reaction process. The embodiment of the present invention chooses to analyze the relevant parameter data in the reactor to perform quality control on potassium dihydrogen phosphate.

[0050] Specifically, in order to implement the potassium dihydrogen phosphate quality control method based on big data analysis proposed in this embodiment, it is first necessary to collect quality inspection data and multidimensional reaction data. The specific process is as follows:

[0051] Firstly, potassium dihydrogen phosphate is divided into current potassium dihydrogen phosphate and historical potassium dihydrogen phosphate according to the production time. The pH value of the solution in the reactor, that is, the acid-base value, is obtained through a pH sensor, and the sequence formed by the pH value at several consecutive moments is obtained, which is recorded as pH time series data; the sequence formed by the temperature inside the reactor at several consecutive moments is obtained through a temperature sensor, which is recorded as temperature time series data; the sequence formed by the stirring speed of the stirrer inside the reactor at several consecutive moments is obtained through a speed sensor of the stirrer inside the reactor, which is recorded as stirring speed time series data.

[0052] Then, the historical quality inspection data of potassium dihydrogen phosphate [cd, p, rj, rw, l] and the reaction data of the historically produced potassium dihydrogen phosphate in the reaction stage are obtained, where cd is purity, p is pH value, rj is solubility, rw is thermal stability, and l is particle size.

[0053] The quality inspection data generally include purity analysis, pH value determination, heavy metal content detection, microbial content detection, solubility test, thermal stability test, particle size distribution, etc. In this embodiment, an array formed by the purity, pH value, solubility, thermal stability and particle size of historical potassium dihydrogen phosphate is selected as the quality inspection data of historical potassium dihydrogen phosphate. The parameters selected for the quality inspection data can be adjusted according to actual conditions, and this embodiment does not specifically limit it.

[0054] Among the several batches of potassium dihydrogen phosphate produced historically, each production batch of potassium dihydrogen phosphate has corresponding reaction data and quality inspection data.

[0055] Then, all the acquired reaction data were subjected to Gaussian filtering to improve the data quality.

[0056] So far, several reaction data of current potassium dihydrogen phosphate and historical potassium dihydrogen phosphate, as well as the quality inspection data of historical potassium dihydrogen phosphate are obtained through the above method.

[0057] Step S002: Based on the similarity of the quality inspection data, the corresponding historical potassium dihydrogen phosphate is divided into several clusters, and the approximate reaction cluster of the current potassium dihydrogen phosphate is screened out according to the similarity in the change trend between the reaction data of the current potassium dihydrogen phosphate and the historical potassium dihydrogen phosphate in different clusters.

[0058] It should be noted that when controlling the production quality of potassium dihydrogen phosphate, when analyzing the real-time acquired reaction data in combination with historical data through a recurrent neural network, since the reaction data is time series data, and RNN has a long-term dependency problem when processing long sequences, that is, it is difficult to capture distant time dependencies. In the production process of potassium dihydrogen phosphate, operations and reactions in the past for a long time will have an important impact on the current product quality, and traditional RNNs may find it difficult to effectively capture such long-term dependencies. At the same time, the computational complexity of RNN training and prediction is high, which may not be suitable for application scenarios requiring real-time prediction, especially production process control that requires rapid response and processing of large amounts of data. Therefore, it is necessary to process the input data of the RNN to reduce the impact of long-term dependency problems on the RNN when processing and analyzing long time series data, and improve the ability of real-time analysis, improve prediction efficiency and accuracy, and thus improve the production quality control effect of potassium dihydrogen phosphate.

[0059] Step 2.1, based on the similarity of the quality inspection data, the corresponding historical potassium dihydrogen phosphate is divided into several clusters.

[0060] As an embodiment, the specific division method is:

[0061] According to the Euclidean distance between the quality inspection data of different historical potassium dihydrogen phosphates, all historical potassium dihydrogen phosphates were clustered using the DBSCAN clustering algorithm to obtain several clusters.

[0062] It should be noted that due to the differences in the production quality of different production batches of potassium dihydrogen phosphate, in order to facilitate the subsequent effective adjustment of the parameters in the reactor according to the current potassium dihydrogen phosphate production quality prediction results, this embodiment classifies the historical potassium dihydrogen phosphate to improve the accuracy of the subsequent prediction of the current potassium dihydrogen phosphate production quality.

[0063] Step 2.2, based on the similarity in the change trend between the current potassium dihydrogen phosphate and the reaction data of historical potassium dihydrogen phosphate in different clusters, the approximate reaction cluster of the current potassium dihydrogen phosphate is screened out.

[0064] As an embodiment, a specific method for obtaining an approximate reaction cluster is as follows:

[0065] Step 2.2.1, based on the change trends of the reaction data of the current potassium dihydrogen phosphate and the historical potassium dihydrogen phosphate in each cluster, respectively obtain the trend information sequence of the current potassium dihydrogen phosphate and the trend information sequence of each cluster.

[0066] As an embodiment, the method for acquiring the trend information sequence of each cluster is:

[0067] First, all historical potassium dihydrogen phosphate reaction data contained in any cluster are respectively subjected to STL time series decomposition to obtain corresponding trend item sequences; a two-dimensional rectangular coordinate system is constructed, and the trend item sequences of potassium dihydrogen phosphate reaction data corresponding to all quality inspection data of the same cluster are placed in the same two-dimensional rectangular coordinate system, and all trend item sequences in the two-dimensional rectangular coordinate system are subjected to curve fitting by the least squares method to obtain a fitting curve of each cluster, which is recorded as a reaction trend curve of the cluster, and the reaction trend curve includes fitting curves corresponding to temperature time series data, pH time series data, and stirring speed time series data, respectively;

[0068] Then, obtain several extreme points of any reaction trend curve, divide the reaction trend curve into several curve segments using the extreme points, obtain the time corresponding to the extreme points at both ends of the curve segment, and perform straight line fitting on the curve segments between any adjacent extreme values ​​to obtain the slope and intercept of the corresponding straight line, obtain the array (t1, t2) formed by the corresponding time of the extreme points at both ends of all curve segments of any reaction trend curve, and the array (k, b) formed by the slope and intercept of the fitting straight line corresponding to all curve segments, which are respectively recorded as the time array and straight line parameter array of the curve segment, and obtain the sequence formed by all the time arrays and straight line parameter arrays in the reaction trend curve according to the time sequence of each curve segment in the reaction trend curve, which is recorded as the trend information sequence of the corresponding cluster {[(t1, t2), (k, b)]1, [(t1, t2), (k, b)]2, …, [(t1, t2), (k, b)] S}, where S is the number of curve segments contained in the reaction trend curve corresponding to the cluster.

[0069] As an embodiment, a method for acquiring a trend information sequence of the current potassium dihydrogen phosphate is as follows: performing STL time series decomposition on any reaction data of the potassium dihydrogen phosphate in the current production process, obtaining corresponding trend items and performing curve fitting, obtaining corresponding reaction trend curves, obtaining extreme points of the reaction trend curves and dividing them into a number of curve segments, obtaining a sequence formed by a time array and a straight line parameter array of each of the curve segments, and thereby obtaining a corresponding trend information sequence of the current potassium dihydrogen phosphate {[(t1, t2), (k, b)]′1, [(t1, t2), (k, b)]′2, …, [(t1, t2), (k, b)]′ S′}, wherein S′ is the number of curve segments contained in the reaction trend curve corresponding to the current potassium dihydrogen phosphate.

[0070] It should be noted that if there is no extreme point in the reaction trend curve corresponding to potassium dihydrogen phosphate in the current production process, the reaction trend curve as a whole is regarded as a curve segment.

[0071] Step 2.2.2, analyze the similarity between the trend information sequence of the current potassium dihydrogen phosphate and the historical potassium dihydrogen phosphate in each cluster, obtain the reaction similarity between the current potassium dihydrogen phosphate and each cluster, and take the cluster corresponding to the maximum reaction similarity as the approximate reaction cluster of the current potassium dihydrogen phosphate.

[0072] It should be noted that, since the reaction data in the reactor during the production of potassium dihydrogen phosphate have obvious trend characteristics, in order to effectively analyze the similarities between the current potassium dihydrogen phosphate and the historical potassium dihydrogen phosphate in the production and preparation process according to the changing trend of the reaction data, the embodiment of the present invention separates the trend items in each reaction data by a time series decomposition method, thereby providing a clear data basis for subsequent analysis and modeling; and by placing the trend item sequence of the potassium dihydrogen phosphate reaction data in the same cluster in a two-dimensional coordinate system for curve fitting, it can help capture the trend items of each reaction data of the historical potassium dihydrogen phosphate in the same cluster. In addition, as the concentration of different ions in the solution changes during the production and preparation of potassium dihydrogen phosphate, the pH value of the solution will also change accordingly. Similarly, the solution temperature and the stirring speed of the stirrer also have stage-by-stage change characteristics. Therefore, the embodiment of the present invention uses the extreme points of the reaction trend curve to divide the curve into multiple curve segments, which can analyze the local characteristics of the trend change in more detail, further improving the subsequent similarity analysis of the reaction trend curve with stage-by-stage trend change characteristics, and screening out clusters that are similar to the current reaction process of potassium dihydrogen phosphate from all clusters.

[0073] As an example, the specific calculation method of reaction similarity is:

[0074]

[0075] Among them, f a is the reaction similarity between the current potassium dihydrogen phosphate and the a-th cluster; x ji represents the i-th element of the j-th trend information sequence of the current potassium dihydrogen phosphate, x′ aji represents the ith element of the jth trend information sequence of the ath cluster, J represents the number of trend information sequences, d() represents obtaining the Euclidean distance, I represents the quantity parameter, I″ represents the number of elements in the trend information sequence corresponding to the current potassium dihydrogen phosphate, and I′ a It indicates the number of elements in the trend information sequence corresponding to the a-th cluster, and min() means obtaining the minimum value.

[0076] It should be noted that, since the reaction data includes pH time series data, temperature time series data and stirring speed time series data, and each reaction data corresponds to a reaction trend curve, each reaction data corresponds to a trend information sequence.

[0077] It should be noted that the trend information sequence reflects the change trend information in different local ranges of the corresponding reaction trend curve. However, since the reaction data of the current potassium dihydrogen phosphate is real-time data obtained during the production and preparation process, the reaction data of the current potassium dihydrogen phosphate is incomplete relative to the historical potassium dihydrogen phosphate. However, considering that the change of reaction data of different batches of potassium dihydrogen phosphate in the production process is different, resulting in differences in the number of divided data segments, in order to fully obtain the change characteristics of the reaction process of the current potassium dihydrogen phosphate and each cluster cluster as a whole, the embodiment of the present invention obtains the Euclidean distance d(x) between the current potassium dihydrogen phosphate and each trend information sequence of the cluster cluster. ji ,x′ aji ), to obtain the reaction similarity, which reflects the similarity between the production and preparation process of the current potassium dihydrogen phosphate and the historical potassium dihydrogen phosphate in each cluster.

[0078] So far, the approximate reaction cluster of the current potassium dihydrogen phosphate is obtained by the above method.

[0079] Step S003: using the correlation between the current potassium dihydrogen phosphate and the historical potassium dihydrogen phosphate before the latest moment to supplement the reaction data of the current potassium dihydrogen phosphate after the latest moment, to obtain new reaction data of the current potassium dihydrogen phosphate, using the slope and value of each data point in the new reaction data and the correlation between the new reaction data and the reaction data of the historical potassium dihydrogen phosphate in the approximate reaction cluster, to correct the new reaction data, to obtain the final reaction data of the current potassium dihydrogen phosphate.

[0080] It should be noted that, since the current potassium dihydrogen phosphate is in the reaction process relative to the historical reaction data of potassium dihydrogen phosphate, the current amount of data of potassium dihydrogen phosphate is relatively small, and the reaction data usually used in training the recurrent neural network is the reaction data corresponding to the entire production process of the entire production batch. Therefore, this embodiment chooses to complete the reaction data of the current potassium dihydrogen phosphate to improve the accuracy of subsequent quality prediction using the recurrent neural network.

[0081] Step 3.1, using the correlation between the current potassium dihydrogen phosphate and the historical potassium dihydrogen phosphate before the latest time, the reaction data after the latest time of the current potassium dihydrogen phosphate is supplemented to obtain new reaction data of the current potassium dihydrogen phosphate.

[0082] Specifically, as an embodiment, first, for any reaction data of current potassium dihydrogen phosphate, the latest time of the reaction data is obtained.

[0083] Then, the latest time is used to divide the corresponding reaction data of all historical potassium dihydrogen phosphate in the approximate reaction cluster into two sections, one is the reaction data section before the latest time, and the other is the reaction data section after the latest time.

[0084] Finally, according to the similarity relationship between the reaction data of the current potassium dihydrogen phosphate and different reaction data segments, and combined with the reaction data segments after the latest reaction of all historical potassium dihydrogen phosphates, the reaction data of the current potassium dihydrogen phosphate is completed to obtain new reaction data of the current potassium dihydrogen phosphate, wherein the new reaction data includes new temperature time series data, new pH time series data and new stirring speed time series data.

[0085] As an embodiment, the calculation method of the value corresponding to the data point after the latest time of the new reaction data is:

[0086]

[0087] Among them, L u represents the value of the data point corresponding to the uth moment after the latest moment of the new reaction data of the current potassium dihydrogen phosphate, fmax represents the reaction similarity between the current potassium dihydrogen phosphate and the approximate reaction cluster, pe m represents the Pearson correlation coefficient between the current potassium dihydrogen phosphate reaction data and the reaction data segment before the latest time of the mth reaction data in the approximate reaction cluster, M represents the number of corresponding reaction data in the approximate reaction cluster, c mu Represents the value of the data point corresponding to the uth time point after the latest time point of the mth reaction data in the approximate reaction cluster.

[0088] It should be noted that, since the data completion process considers the similarity relationship between the current reaction data of potassium dihydrogen phosphate and different reaction data segments, but does not take into account the mutual influence relationship between different reaction data of potassium dihydrogen phosphate in the reaction stage, the embodiment of the present invention corrects the completed data through the mutual influence relationship between different reaction data.

[0089] Step 3.2, using the slope and value of each data point in the new reaction data and the correlation between the new reaction data and the reaction data of historical potassium dihydrogen phosphate in the approximate reaction cluster, the new reaction data is corrected to obtain the final reaction data of the current potassium dihydrogen phosphate.

[0090] Firstly, the correction coefficient of each data point in the new reaction data is obtained according to the slope and size of each data point in any new reaction data and the correlation between the new reaction data and the reaction data of historical potassium dihydrogen phosphate in the approximate reaction cluster.

[0091] As an embodiment, the specific method for obtaining the correction coefficient of each data point in the new reaction data is:

[0092]

[0093] Among them, α1 represents the new temperature time series data, α2 represents the new pH time series data, α3 represents the new stirring speed time series data, γ(α u ) represents the correction coefficient of the u-th data point in the new reaction data α, represents the approximation factor of the new reaction data α, δ u () represents the numerical parameter of the uth data point of the new reaction data, δ1′ u represents the appropriate numerical parameter of the uth data point in the new temperature time series data, || represents obtaining the absolute value, norm() represents the normalization function, and exp() represents the exponential function with a natural constant as the base.

[0094] The numerical parameter is the product of the slope of the data point in the new reaction data and the numerical value.

[0095] The suitable numerical parameter is the product of the slope of the data point in the new temperature time series data and Br, where Br is the suitable temperature of potassium dihydrogen phosphate in the reactor during the reaction process.

[0096] It should be noted that the normalization range of norm() in this embodiment is (A, B), where A and B are the preset first parameter and second parameter. The first parameter A=0.5 and the second parameter B=1.5 are preset based on experience. They can be adjusted according to actual conditions and are not specifically limited in this embodiment.

[0097] It should be noted that the production of potassium dihydrogen phosphate usually involves an acid-base neutralization reaction, and under acidic conditions (lower pH values), the acid-base neutralization reaction may proceed faster because the addition of potassium hydroxide (KOH) or other alkaline substances can more quickly neutralize the phosphate ions (PO4 3- ), therefore, the smaller the pH value of the reaction solution, the faster the reaction proceeds. At the same time, because the neutralization reaction is an exothermic reaction, the more heat is released, the temperature of the solution in the reactor will rise rapidly. Conversely, under alkaline conditions (higher pH value), the neutralization may be slower, causing the temperature of the solution in the reactor to change slowly.

[0098] In addition, the faster the stirring speed of the stirrer in the reactor, the more uniform the different ions in the reactor are in the solution, the faster the acid-base neutralization reaction is, the more heat is released by the reaction, and the higher the temperature of the solution is. The smaller the difference between the temperature of the solution and the theoretical temperature is, the greater the |δ u (α1)-δ1′ u| is smaller, so that the speed of the reaction in the solution will be accelerated, that is, the neutralization degree of acid and alkali in the solution is higher, and the pH value of the solution gradually shifts from acidic to neutral. Therefore, this embodiment takes into account the influence relationship between various factors in the production and preparation process of potassium dihydrogen phosphate. In order to make the new reaction data closer to the actual situation, according to the slope and size of each data point in the new reaction data and the correlation between the new reaction data and the reaction data of historical potassium dihydrogen phosphate in the approximate reaction cluster, the correction coefficient of each new reaction data is obtained, so as to correct the new reaction data and obtain more accurate final reaction data of the current potassium dihydrogen phosphate.

[0099] Then, the new reaction data is corrected using the correction coefficient to obtain the final reaction data of the current potassium dihydrogen phosphate.

[0100] As an example, the specific method for obtaining the final reaction data is as follows:

[0101] L′(α u )=γ(α u )×L(α u )

[0102] Among them, L(α′ u ) represents the value of the data point corresponding to the uth time after the latest time in the final reaction data α′, γ(α u ) represents the correction coefficient of the data point corresponding to the uth time after the latest time in the new reaction data α, L(α u ) represents the value of the data point corresponding to the uth moment after the latest moment in the new reaction data α.

[0103] At this point, the final reaction data of the current potassium dihydrogen phosphate is obtained through the above method.

[0104] Step S004: based on the distribution of the quality inspection data in each cluster, obtain representative quality inspection data of each cluster.

[0105] It should be noted that the embodiment of the present invention takes into account that the amount of historical potassium dihydrogen phosphate data is very large. When the quality inspection data of historical potassium dihydrogen phosphate is used as a label to train the RNN, the RNN will try to remember the specific features of each label instead of learning the general rules between them, resulting in longer training time and increased prediction time. In addition, too many label categories may introduce noise, such as confusion or misclassification between labels, which will affect the learning effect of the model. Therefore, in order to avoid the above problems, the embodiment of the present invention performs weighted fusion on the quality inspection data of historical potassium dihydrogen phosphate belonging to the same cluster to obtain more representative quality inspection data and reduce the number of labels.

[0106] Specifically, for any cluster, the Euclidean distance between each quality inspection data in the cluster and the cluster center of the cluster is obtained, and all quality inspection data in the cluster are weightedly fused using the Euclidean distance to obtain representative quality inspection data of the cluster.

[0107] As an embodiment, the calculation method of the representative quality inspection data of any cluster is:

[0108]

[0109] Among them, D′ represents the quality inspection data, D0 represents the mode of all quality inspection data in the cluster, and D n represents the nth quality inspection data in the cluster, d′ n It represents the Euclidean distance between the nth quality inspection data in the cluster and the cluster center of the cluster to which it belongs. g and r are the preset third and fourth parameters respectively. exp() represents an exponential function with a natural constant as the base.

[0110] It should be noted that the third parameter g and the fourth parameter r are preset to be 0.5 and 2 respectively based on experience, and can be adjusted according to actual conditions, and are not specifically limited in this embodiment.

[0111] So far, the representative quality inspection data of each cluster is obtained through the above method.

[0112] Step S005: Combine the representative quality inspection data and use the neural network to predict the quality of the current potassium dihydrogen phosphate, and adjust the parameters in the current potassium dihydrogen phosphate production process through the decision tree to achieve quality control of the current potassium dihydrogen phosphate.

[0113] Step 5.1, combining representative quality inspection data and using a neural network to predict the quality of the current potassium dihydrogen phosphate.

[0114] Specifically, first, a recurrent neural network is constructed.

[0115] Then, the trend information sequences corresponding to all the historical potassium dihydrogen phosphate reaction data are obtained respectively, a trend information sequence of a historical potassium dihydrogen phosphate is taken as a sample, and the representative quality inspection data corresponding to the cluster to which the historical potassium dihydrogen phosphate belongs is taken as the label of the corresponding sample, and a set formed by all the labeled samples is obtained as a data set for training the recurrent neural network, and the data set is divided into a training set, a validation set, and a test set in a ratio of 6:3:1. The data set is used to train the recurrent neural network according to the cross entropy loss function and the stochastic gradient descent algorithm to converge, and a trained recurrent neural network is obtained as a quality prediction neural network.

[0116] Finally, the trend information sequence of the final reaction data of the current potassium dihydrogen phosphate is obtained, and the trend information sequence is used as the input of the quality prediction neural network, and the quality prediction neural network is used to perform analysis to obtain the predicted quality data of the current potassium dihydrogen phosphate.

[0117] Step 5.2, adjusting the parameters in the current potassium dihydrogen phosphate production process through the decision tree to achieve quality control of the current potassium dihydrogen phosphate.

[0118] As an embodiment, the specific control method is: using a decision tree to analyze the predicted quality data, and adjusting and controlling the temperature, pH and speed of the stirrer in the reactor during the production process of potassium dihydrogen phosphate according to the output of the decision tree, thereby achieving quality control of potassium dihydrogen phosphate.

[0119] like Figure 2 Shown is a schematic diagram of the quality control process of potassium dihydrogen phosphate.

[0120] At this point, this embodiment is completed.

[0121] It should be noted that the exp(-x) model used in this embodiment is only used to indicate that the negative correlation and the result of the constraint model output are in the interval (0,1). In specific implementation, it can be replaced with other models with the same purpose. This embodiment only uses the exp(-x) model as an example for description without making specific limitations on it, where x refers to the input of the model.

[0122] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. Potassium dihydrogen phosphate quality control method based on big data analysis, It is characterized in that The method comprises the following steps: Obtain some reaction data of current and historical potassium dihydrogen phosphate, as well as historical quality inspection data of potassium dihydrogen phosphate; Based on the similarity of the quality inspection data, the corresponding historical potassium dihydrogen phosphate is divided into several clusters, and the approximate reaction clusters of the current potassium dihydrogen phosphate are screened out according to the similarity in the change trend between the reaction data of the current potassium dihydrogen phosphate and the historical potassium dihydrogen phosphate in different clusters; The reaction data after the latest moment of the current potassium dihydrogen phosphate is supplemented by using the correlation between the current potassium dihydrogen phosphate and the historical potassium dihydrogen phosphate before the latest moment, so as to obtain the new reaction data of the current potassium dihydrogen phosphate, and the new reaction data is corrected by using the slope and value of each data point in the new reaction data and the correlation between the new reaction data and the reaction data of the historical potassium dihydrogen phosphate in the approximate reaction cluster, so as to obtain the final reaction data of the current potassium dihydrogen phosphate; Based on the distribution of quality inspection data in each cluster, obtain representative quality inspection data of each cluster; The quality of the current potassium dihydrogen phosphate is predicted by combining representative quality inspection data and using a neural network. The parameters in the current potassium dihydrogen phosphate production process are adjusted through a decision tree to achieve quality control of the current potassium dihydrogen phosphate.

2. The potassium dihydrogen phosphate quality control method based on big data analysis according to claim 1, It is characterized in that The method of screening out the approximate reaction cluster of the current potassium dihydrogen phosphate according to the similarity in the change trend between the reaction data of the current potassium dihydrogen phosphate and the historical potassium dihydrogen phosphate in different clusters includes the following specific methods: Based on the change trends of the reaction data of the current potassium dihydrogen phosphate and the historical potassium dihydrogen phosphate in each cluster, the trend information sequence of the current potassium dihydrogen phosphate and the trend information sequence of each cluster are obtained respectively; The similarity between the trend information sequence of the current potassium dihydrogen phosphate and the historical potassium dihydrogen phosphate in each cluster is analyzed to obtain the reaction similarity between the current potassium dihydrogen phosphate and each cluster, and the cluster corresponding to the maximum reaction similarity is taken as the approximate reaction cluster of the current potassium dihydrogen phosphate.

3. According to the potassium dihydrogen phosphate quality control method based on big data analysis according to claim 2, It is characterized in that The specific method for obtaining the trend information sequence of the cluster is: Perform STL time series decomposition on all historical potassium dihydrogen phosphate reaction data contained in any cluster to obtain the corresponding trend item sequence; construct a two-dimensional rectangular coordinate system, place the potassium dihydrogen phosphate reaction data trend item sequence corresponding to all quality inspection data of the same cluster into the same two-dimensional rectangular coordinate system, use the least squares method to perform curve fitting on all trend item sequences in the two-dimensional rectangular coordinate system, and obtain the fitting curve of each cluster, which is recorded as the reaction trend curve of the cluster; Obtain several extreme points of any reaction trend curve, divide the reaction trend curve into several curve segments using the extreme points, obtain the time corresponding to the extreme points at both ends of the curve segment, and perform straight line fitting on the curve segments between any adjacent extreme values ​​to obtain the slope and intercept of the corresponding straight line, obtain an array formed by the corresponding time of the extreme points at both ends of all curve segments of any reaction trend curve, and an array formed by the slope and intercept of the corresponding fitting straight line of all curve segments, which are respectively recorded as the time array and the straight line parameter array of the curve segment, and obtain a sequence formed by all the time arrays and the straight line parameter arrays in the reaction trend curve according to the time sequence of each curve segment in the reaction trend curve, which is recorded as the trend information sequence of the corresponding cluster cluster.

4. The potassium dihydrogen phosphate quality control method based on big data analysis according to claim 3, It is characterized in that The specific method for obtaining the trend information sequence of the current potassium dihydrogen phosphate is: Perform STL time series decomposition on any reaction data of potassium dihydrogen phosphate in the current production process, obtain the corresponding trend item and perform curve fitting, obtain the corresponding reaction trend curve, obtain the extreme points of the reaction trend curve and divide it into several curve segments, obtain the sequence formed by the time array and the straight line parameter array of each curve segment, so as to obtain the corresponding trend information sequence of the current potassium dihydrogen phosphate.

5. The potassium dihydrogen phosphate quality control method based on big data analysis according to claim 2, It is characterized in that The specific calculation method of the reaction similarity is: Among them, f a is the reaction similarity between the current potassium dihydrogen phosphate and the a-th cluster; x ji represents the i-th element of the j-th trend information sequence of the current potassium dihydrogen phosphate, x ′ aji represents the ith element of the jth trend information sequence of the ath cluster, J represents the number of trend information sequences, d() represents obtaining the Euclidean distance, I represents the quantity parameter, I″ represents the number of elements of the trend information sequence corresponding to the current potassium dihydrogen phosphate, and I ′ a It indicates the number of elements in the trend information sequence corresponding to the a-th cluster, and min() means obtaining the minimum value.

6. The potassium dihydrogen phosphate quality control method based on big data analysis according to claim 1, It is characterized in that The specific method for obtaining the new reaction data of the current potassium dihydrogen phosphate is: For any reaction data of current potassium dihydrogen phosphate, obtain the latest time of the reaction data; Using the latest time, the corresponding reaction data of all historical potassium dihydrogen phosphate in the approximate reaction cluster are divided into two sections, one is the reaction data section before the latest time, and the other is the reaction data section after the latest time; According to the similarity relationship between the reaction data of the current potassium dihydrogen phosphate and different reaction data segments, and in combination with the reaction data segments after the latest reaction of all historical potassium dihydrogen phosphates, the reaction data of the current potassium dihydrogen phosphate is completed to obtain new reaction data of the current potassium dihydrogen phosphate, wherein the new reaction data includes new temperature time series data, new pH time series data and new stirring speed time series data.

7. The potassium dihydrogen phosphate quality control method based on big data analysis according to claim 1, It is characterized in that The specific method for obtaining the final reaction data of the current potassium dihydrogen phosphate is: Obtaining a correction coefficient for each data point in the new reaction data according to the slope and size of each data point in any new reaction data and the correlation between the new reaction data and the reaction data of historical potassium dihydrogen phosphate in the approximate reaction cluster; The new reaction data is corrected using the correction coefficient to obtain the final reaction data of the current potassium dihydrogen phosphate.

8. The potassium dihydrogen phosphate quality control method based on big data analysis according to claim 7, It is characterized in that The specific calculation method of the correction coefficient of each data point in the new reaction data is: Among them, α1 represents the new temperature time series data, α2 represents the new pH time series data, α3 represents the new stirring speed time series data, γ(α u ) represents the correction coefficient of the u-th data point in the new reaction data α, represents the approximation factor of the new reaction data α, δ u () represents the numerical parameter of the uth data point of the new reaction data, δ1 ′ u represents the appropriate numerical parameter of the u-th data point in the new temperature time series data, || represents obtaining the absolute value, norm() represents the normalization function, exp() represents the exponential function with the natural constant as the base; the numerical parameter is the product of the slope of the data point in the new reaction data and the numerical value; the appropriate numerical parameter is the product of the slope of the data point in the new temperature time series data and Br, and Br is the appropriate temperature of potassium dihydrogen phosphate in the reactor during the reaction process.

9. The potassium dihydrogen phosphate quality control method based on big data analysis according to claim 1, It is characterized in that The method of obtaining representative quality inspection data of each cluster based on the distribution of quality inspection data in each cluster includes: For any cluster, the Euclidean distance between each quality inspection data in the cluster and the cluster center of the cluster is obtained, and all quality inspection data in the cluster are weightedly fused using the Euclidean distance to obtain representative quality inspection data of the cluster.

10. The potassium dihydrogen phosphate quality control method based on big data analysis according to claim 7, It is characterized in that The method of using the correction coefficient to correct the new reaction data to obtain the final reaction data of the current potassium dihydrogen phosphate includes: The calculation method for the value of the data point after the latest time in the final reaction data is: L ′ (a u )=γ(α u )×L(a u ) Among them, L(α ′ u ) represents the final reaction data α ′ The value of the data point corresponding to the uth moment after the latest moment in , γ(α u ) represents the correction coefficient of the data point corresponding to the uth time after the latest time in the new reaction data α, L(α u ) represents the value of the data point corresponding to the uth moment after the latest moment in the new reaction data α.

Citation Information

Patent Citations

  • Self-adaptive temperature early warning method and system based on hydrocracking unit

    CN116227673A

  • Machine learning-based quality control of a culture for bioproduction

    US20220382266A1

  • Method and system for optimizing an alkane dehydrogenation operation

    US20240210917A1

Cited By

  • High-quality phosphate chemical process automation system

    CN120295253A