A method and system for monitoring the operation status of a crystallizer

Abstract

The present invention relates to the field of data processing technology, and more specifically, to a method and system for monitoring the operation status of a crystallizer, the method comprising: constructing a process memory matrix of the crystallizer, using the ratio of the cumulative sum of correlation coefficients of any two data in the current crystallization stage and the next crystallization stage, predicting the predicted differential vector of the next crystallization stage, using the column vector corresponding to the next crystallization stage in the process memory matrix as a reference vector, based on the distance between the differential vector of the next crystallization stage and the same data in the reference vector, measuring the similarity of the corresponding two vectors, calculating the abnormality of the next crystallization stage based on the similarity and the cosine similarity between the differential vector of the next crystallization stage and the predicted differential vector, and using the comparison result of the abnormality and a preset threshold value to monitor the state of the crystallizer. The present invention can realize accurate monitoring of the operation status of the crystallizer.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and more specifically, to a method and system for monitoring the operation status of a crystallizer. Background Art

[0002] A crystallizer is a device used in industrial production to separate solid crystals from a solution, melt or gas through a crystallization process. Monitoring the operating status of the crystallizer is essential to ensure product quality, improve equipment reliability, and ensure production safety. It also provides valuable data support for process optimization. Therefore, it is crucial to accurately monitor the operating status of the crystallizer.

[0003] In the related art, a method for intelligently monitoring the operating status of electrical equipment is disclosed in a patent application document with publication number CN116610482A. The method includes: obtaining temperature change data information of the electrical equipment and constructing a time-series temperature change characteristic curve, obtaining abnormal temperature data in the time-series temperature change characteristic curve, obtaining a first non-mutation degree and a second non-mutation degree of the abnormal temperature data, obtaining a final non-mutation degree based on the first non-mutation degree and the second non-mutation degree, and evaluating the impact value of abnormal data points with a higher non-mutation degree, so as to process the abnormal temperature data of the electrical equipment according to the impact value and achieve the effect of intelligently monitoring abnormal data.

[0004] However, in the above scheme, the monitoring of the equipment operating status only relies on a single parameter (i.e., temperature change data information). Although the temperature change data information can reflect the working status of the equipment to a certain extent, evaluating the equipment operating status based on data changes in a single dimension may not fully reflect its overall health status and lack sufficient diagnostic capabilities. This will result in low accuracy in identifying abnormal operating conditions, thereby affecting the accuracy of equipment operating status monitoring. Summary of the invention

[0005] In order to solve the problem that the operating status of a device cannot be accurately monitored, the present invention provides a method and system for monitoring the operating status of a crystallizer.

[0006] According to a first aspect of the present invention, there is provided a method for monitoring a crystallizer operation state, comprising:

[0007] Based on various types of intact data pre-acquired at different crystallization stages, a process memory matrix of the crystallizer is constructed, in which the row vector represents the change rate of any type of intact data at each crystallization stage, and the column vector represents the change rate of each type of intact data at any crystallization stage;

[0008] A one-dimensional vector consisting of the difference between the current moment and the next moment of each data in the current crystallization period is used as the differential vector of the current crystallization stage, and the ratio of the cumulative sum of the correlation coefficients of any two data in the current crystallization stage and the next crystallization stage is calculated. The differential vector is weighted by the normalized value of the absolute value of the ratio to obtain the predicted differential vector of the next crystallization stage;

[0009] The column vector corresponding to the next crystallization stage in the process memory matrix is ​​used as the reference vector. Based on the distance between the difference vector of the next crystallization stage and the same data in the reference vector, the similarity between the two vectors is measured. The similarity is negatively correlated with the distance.

[0010] The abnormality degree of the next crystallization stage is calculated. The abnormality degree is negatively correlated with the similarity degree, and the absolute value of the cosine similarity between the differential vector of the next crystallization stage and the predicted differential vector is negatively correlated. The state of the crystallizer is monitored based on the comparison result between the abnormality degree and the preset threshold.

[0011] The present invention overcomes the limitations of a single data monitoring device in monitoring the operating status of the crystallizer. By combining multiple data sources to achieve comprehensive monitoring of the operating status of the crystallizer, the present invention can more accurately reflect the overall operating conditions of the crystallizer, thereby improving the accuracy of monitoring the operating status of the crystallizer and ensuring accurate monitoring of the operating status of the crystallizer.

[0012] Preferably, the data in any crystallization stage includes temperature data, vibration data and pressure data.

[0013] The data types selected by the present invention are key physical parameters in the operation process of the crystallizer, which can comprehensively reflect the status of the crystallizer equipment from multiple angles and provide sufficient data support for accurately monitoring the operating status of the crystallizer.

[0014] Preferably, the normalized value is used as the data change synchronization rate between the current crystallization stage and the next crystallization stage. The data change synchronization rate satisfies the following relationship:

[0015] ;

[0016] In the formula, For the The crystallization stage and The data change synchronization rate of each crystallization stage; , Respectively The crystallization stage and Correlation coefficient between temperature data and pressure data in each crystallization stage; , Respectively The crystallization stage and Correlation coefficient between temperature data and vibration data in each crystallization stage; , Respectively The crystallization stage and Correlation coefficient between pressure data and vibration data in each crystallization stage; is the absolute value symbol; is the normalization function.

[0017] The data change synchronization rate calculated by the present invention can measure the possibility of abnormality in the operation state of the crystallizer.

[0018] Preferably, the correlation coefficient is the Pearson correlation coefficient or the Spearman rank correlation coefficient.

[0019] Preferably, the method for obtaining the rate of change includes:

[0020] A differential sequence of any kind of intact data in any crystallization stage is obtained, and the average value of the data in the differential sequence is used as the change rate of the intact data in the crystallization stage.

[0021] The present invention constructs a memory process matrix of the crystallizer based on the change rates of various types of intact data in different crystallization stages, which can comprehensively reflect the data changes under normal operating conditions in different crystallization stages.

[0022] Preferably, the abnormality degree satisfies the following relationship:

[0023] ;

[0024] In the formula, For the The degree of abnormality of each crystallization stage; For the The similarity between the difference vector of each crystallization stage and the reference vector; For the The difference vector of the crystallization stage; For the The predicted difference vector of each crystallization stage; is the representation of cosine similarity; is the natural exponential function; is the absolute value symbol.

[0025] The present invention combines a variety of data to determine the degree of abnormality, thereby ensuring the accuracy of the degree of abnormality.

[0026] Preferably, the similarity satisfies the following relationship:

[0027] ;

[0028] In the formula, For the The similarity between the difference vector of each crystallization stage and the reference vector; For the The difference vector of the crystallization stage elements; For the The reference vector of the crystallization stage elements; for the summation symbol; is the normalization function.

[0029] Preferably, monitoring the state of the crystallizer based on the comparison result of the abnormality degree and the preset threshold value includes:

[0030] Obtaining a preset threshold; if the degree of abnormality in the next crystallization stage of the current crystallization stage is greater than or equal to the preset threshold, determining that the operation state of the crystallizer in the next crystallization stage is abnormal;

[0031] If it is less than the preset threshold, it is determined that the operation status of the crystallizer in the next crystallization stage is normal.

[0032] According to a second aspect of the present invention, a crystallizer operation status monitoring system is provided. The system includes a memory and a processor. The memory stores a computer program. The processor executes the computer program to implement the steps of the first aspect of the present invention.

[0033] The present invention has the following effects:

[0034] 1. The present invention overcomes the limitations of a single data monitoring device in operating status monitoring. By combining multiple data sources, multiple parameters that can effectively evaluate the possibility of abnormal operating status of the crystallizer are calculated, thereby achieving accurate monitoring of the operating status of the crystallizer.

[0035] 2. The present invention constructs a process memory matrix of the crystallizer to comprehensively reflect the changes of the crystallizer under normal operating conditions in each crystallization stage, thereby providing data support for evaluating the possibility of abnormal operating conditions of the crystallizer.

[0036] 3. The present invention combines various data to calculate the degree of abnormality, thereby ensuring the accuracy of the degree of abnormality, thereby enabling accurate monitoring of the operating status of the crystallizer based on the magnitude of the degree of abnormality in the corresponding crystallization stage. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] The above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily understood by reading the detailed description below with reference to the accompanying drawings. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein:

[0038] Figure 1 It is a schematic flow chart of the steps of a method for monitoring the operating status of a crystallizer according to an embodiment of the present invention. DETAILED DESCRIPTION

[0039] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0040] The specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0041] Reference Figure 1 , a method for monitoring the operation status of a crystallizer, comprising steps S1 to S4, specifically as follows:

[0042] S1: Based on various types of intact data pre-acquired in different crystallization stages, a process memory matrix of the crystallizer is constructed. The row vectors in the process memory matrix represent the change rate of any type of intact data in each crystallization stage, and the column vectors represent the change rate of each type of intact data in any crystallization stage.

[0043] In an exemplary embodiment of the present invention, good data refers to data without anomalies, noise or errors. By constructing a process memory matrix of a crystallizer based on various types of good data in different crystallization stages, the normal operation status of various types of data in the historical crystallization stage can be accurately reflected, thereby providing a data basis for monitoring the abnormal operation status of the crystallizer.

[0044] In an exemplary embodiment of the present invention, the data in any crystallization stage includes temperature data, vibration data, and pressure data.

[0045] Specifically, the readings of the temperature sensor, vibration sensor and pressure sensor electrically connected to the crystallizer can be collected at a certain frequency, such as 1 time / second, in each crystallization stage to obtain the temperature data sequence, vibration data sequence and pressure data sequence in the corresponding crystallization stage, that is, various types of data corresponding to the crystallization stage.

[0046] Furthermore, after obtaining multiple data sequences of each crystallization stage, all data sequences can be preprocessed to achieve denoising, eliminate erroneous data, and remove abnormal data, so as to obtain multiple types of intact data in different crystallization stages, thereby ensuring that accurate data is used when constructing the process memory matrix of the crystallizer; then, the change rate of any type of intact data in any crystallization stage can be calculated to construct the process memory matrix of the crystallizer.

[0047] In an exemplary embodiment of the present invention, the determination of the change rate of any good data in any crystallization stage can be achieved by the following steps:

[0048] A differential sequence of any kind of intact data at any crystallization stage is obtained, and the average value of the data in the differential sequence is used as the change rate of the intact data at the crystallization stage.

[0049] Specifically, the change rate of any intact data in any crystallization stage satisfies the following relationship:

[0050] ;

[0051] In the formula, For the In the crystallization stage The rate of change of good data; For the The crystallization stage, The first time collected The value of the intact data; For the The crystallization stage, The first time collected The value of the intact data; For the The total number of sampling moments within a crystallization stage; is the summation symbol.

[0052] in, Reflected in the The first The first-order difference of the intact data of this type; the sequence composed of the first-order differences of the intact data of this type collected at all adjacent moments in the crystallization stage is the difference sequence of the intact data of this type in the crystallization stage.

[0053] In another embodiment, the change rate of any intact data in any crystallization stage can also be determined using a traditional change rate calculation formula, that is, determined using the ratio of the data change amount and time. This embodiment does not specifically limit the method for determining the change rate.

[0054] Furthermore, after determining the change rate of each intact data in each crystallization stage, the process memory matrix of the crystallizer can be constructed. Exemplarily, the process memory matrix of the crystallizer can be expressed as:

[0055] ;

[0056] in, is the process memory matrix of the crystallizer; , as well as They are the change rates of intact temperature data, intact pressure data and intact vibration data in the first crystallization stage respectively; , as well as They are the change rates of intact temperature data, intact pressure data and intact vibration data in the second crystallization stage respectively; , as well as Respectively The rate of change of intact temperature data, intact pressure data and intact vibration data within each crystallization stage; is the crystallization stage number.

[0057] Optionally, based on the change rate of each intact data in each crystallization stage, constructing a process memory matrix of the crystallizer can comprehensively reflect the data changes of the crystallizer under normal operating conditions in different crystallization stages, thereby providing an effective data basis for subsequent analysis.

[0058] S2: Take the one-dimensional vector formed by the difference between the current moment and the next moment of each data in the current crystallization period as the differential vector of the current crystallization stage, calculate the ratio of the cumulative sum of the correlation coefficients of any two data in the current crystallization stage and the next crystallization stage, and use the normalized value of the absolute value of the ratio to weight the differential vector to obtain the predicted differential vector of the next crystallization stage.

[0059] For example, for The difference between the temperature data at the current moment and the next moment in this stage can be recorded as ; The difference between the pressure data at the current moment and the next moment in this stage is recorded as ; The difference between the vibration data at the current moment and the next moment in this stage is recorded as ; then , as well as One-dimensional vector composed of , recorded as the difference vector of this crystallization stage .

[0060] In an exemplary embodiment of the present invention, the correlation coefficient of any two data in the current crystallization stage and the next crystallization stage can be a Pearson correlation coefficient or a Spearman rank correlation coefficient. It should be noted that the Pearson correlation coefficient is used in the present invention. Among them, the determination method of the Pearson correlation coefficient between the data is a prior art, and this embodiment will not be described in detail here.

[0061] Furthermore, after the correlation coefficients of any two data in the corresponding two crystallization stages are determined, the ratio of the cumulative sum of the Pearson correlation coefficients of any two data in the current crystallization stage to the cumulative sum of the Pearson correlation coefficients of any two data in the next crystallization stage can be calculated.

[0062] In an exemplary embodiment of the present invention, the normalized value of the absolute value of the ratio is used as the data change synchronization rate between the current crystallization stage and the next crystallization stage. Specifically, the data change synchronization rate between the current crystallization stage and the next crystallization stage satisfies the following relationship:

[0063] ;

[0064] In the formula, For the The crystallization stage and The data change synchronization rate of each crystallization stage; , Respectively The crystallization stage and Correlation coefficient between temperature data and pressure data in each crystallization stage; , Respectively The crystallization stage and Correlation coefficient between temperature data and vibration data in each crystallization stage; , Respectively The crystallization stage and Correlation coefficient between pressure data and vibration data in each crystallization stage; is the absolute value symbol; is the normalization function.

[0065] in, The larger the value is, the higher the synchronous change rate of the data in the corresponding two crystallization stages is, and the smaller the possibility of abnormal operation of the crystallizer is; on the contrary, the smaller the value is, the lower the synchronous change rate of the data in the corresponding two crystallization stages is, and the greater the possibility of abnormal operation of the crystallizer is.

[0066] Furthermore, the data change synchronization rate of the current crystallization stage and the next crystallization stage can be used to predict the differential vector of the next crystallization stage to obtain the predicted differential vector of the next crystallization stage. It should be noted that the differential vector of the next crystallization stage is determined in the same way as the differential vector of the current crystallization stage, and this embodiment will not be described in detail here.

[0067] Specifically, the data change synchronization rate between the current crystallization stage and the next crystallization stage may be used as a weight, and then the differential vector of the current crystallization stage may be weighted to obtain a predicted differential vector of the next crystallization stage.

[0068] For example, if The difference vector of the crystallization stage is , No. The crystallization stage and The data change synchronization rate of the crystallization stage is When The predicted difference vector of the crystallization stage : ; In the formula, For the The rate of change of temperature data in each crystallization stage; For the The rate of change of pressure data in each crystallization stage; For the The rate of change of vibration data in each crystallization stage; For the The crystallization stage and The data change synchronization rate of each crystallization stage; A symbol for representing a vector.

[0069] S3: The column vector corresponding to the next crystallization stage in the process memory matrix is ​​used as a reference vector. Based on the distance between the differential vector of the next crystallization stage and the same data in the reference vector, the similarity between the two corresponding vectors is measured. The similarity is negatively correlated with the distance.

[0070] For example, if the next crystallization stage of the current crystallization stage is crystallization stage, the reference vector is the column vector corresponding to this stage in the process memory matrix, that is, A column vector.

[0071] Specifically, the degree of similarity between the difference vector of the next crystallization stage of the current crystallization stage and the reference vector satisfies the following relationship:

[0072] ;

[0073] In the formula, For the The similarity between the difference vector of each crystallization stage and the reference vector; For the The difference vector of the crystallization stage elements; For the The reference vector of the crystallization stage elements; for the summation symbol; is the normalization function.

[0074] in, Reflects the The first difference vector of the crystallization stage element, and the reference vector of this vector The distance between the elements can be the Euclidean distance or the Manhattan distance. This embodiment does not specifically limit the distance used to measure the similarity between the two corresponding vectors.

[0075] Reflects the The sum of the distances between the differential vector of each crystallization stage and the same data in the reference vector of the vector. The larger the value, the lower the similarity between the two corresponding vectors, and the possibility of abnormal operation of the crystallizer is relatively large; on the contrary, the smaller the value, the higher the similarity between the two corresponding vectors, and the possibility of abnormal operation of the crystallizer is relatively small.

[0076] In another embodiment, the similarity between the two corresponding vectors may also be calculated using other formulas, such as not using a ratio calculation formula but using a negation calculation formula.

[0077] S4: Calculate the abnormality level of the next crystallization stage. The abnormality level is negatively correlated with the similarity level, and the absolute value of the cosine similarity between the differential vector of the next crystallization stage and the predicted differential vector; monitor the state of the crystallizer based on the comparison result between the abnormality level and the preset threshold.

[0078] Specifically, the abnormality degree of the next crystallization stage of the current crystallization stage satisfies the following relationship:

[0079] ;

[0080] In the formula, For the The degree of abnormality of each crystallization stage; For the The similarity between the difference vector of each crystallization stage and the reference vector; For the The difference vector of the crystallization stage; For the The predicted difference vector of each crystallization stage; is the representation of cosine similarity; is the natural exponential function; is the absolute value symbol.

[0081] Among them, since the value range of the cosine similarity between vectors is (-1, 1), the present invention adjusts the value range to (0, 1) by taking the absolute value, thereby simplifying the value range of the data and reducing the difficulty of calculation.

[0082] Optional, when When it is smaller, the The difference vector of the first crystallization stage is less similar to the reference vector, which indicates that the possibility of abnormal operation of the crystallizer in this crystallization stage is relatively high. The abnormality of the crystallization stage is relatively large; at this time, if When it is also small, it means that the first The more the predicted differential vector of a crystallization stage is different from the actual differential vector of the crystallization stage, the more likely the operation state of the crystallizer in the crystallization stage is abnormal. The degree of abnormality in each crystallization stage is relatively large.

[0083] Further, after determining the degree of abnormality of the crystallizer in the next crystallization stage of the current crystallization stage, the operating state of the crystallizer can be monitored based on the degree of abnormality.

[0084] In an exemplary embodiment of the present invention, the operating status of the crystallizer can be monitored by the following steps:

[0085] Obtain a preset threshold; if the degree of abnormality in the next crystallization stage of the current crystallization stage is greater than or equal to the preset threshold, it is determined that the operating state of the crystallizer in the next crystallization stage is abnormal; if it is less than the preset threshold, it is determined that the operating state of the crystallizer in the next crystallization stage is normal.

[0086] Optionally, the threshold can be set to 0.5. If the abnormality degree of the first crystallization stage is greater than or equal to 0.5, it is determined that the operation state of the crystallizer in this stage is abnormal; if the abnormality degree of the first crystallization stage is greater than or equal to 0.5, it is determined that the operation state of the crystallizer in this stage is abnormal; If the abnormality degree of each crystallization stage is less than 0.5, it is determined that the operation state of the crystallizer in this stage is normal, thereby realizing the monitoring of the operation state of the crystallizer.

[0087] The present invention also provides a crystallizer operation status monitoring system, the system includes a memory and a processor, and a computer program is stored in the memory, the computer program integrates the function of a crystallizer operation status monitoring method, when the computer program is executed, the crystallizer operation status can be accurately monitored through a crystallizer operation status monitoring method.

[0088] In the description of this specification, "plurality" or "several" means at least two, such as two, three or more, etc., unless otherwise clearly and specifically defined.

[0089] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.

Claims

1. A method for monitoring the operation status of a crystallizer, characterized in that: include: Based on multiple types of intact data pre-acquired at different crystallization stages, a process memory matrix of the crystallizer is constructed, wherein the row vectors in the process memory matrix represent the change rate of any type of intact data at each crystallization stage, and the column vectors represent the change rate of each type of intact data at any crystallization stage; The data in any crystallization stage include temperature data, vibration data and pressure data; The method for obtaining the change rate includes: obtaining a differential sequence of any kind of intact data in any crystallization stage, and taking the average value of the data in the differential sequence as the change rate of the intact data in the crystallization stage; A one-dimensional vector consisting of the difference between the current moment and the next moment of each data in the current crystallization period is used as the differential vector of the current crystallization stage, and the ratio of the cumulative sum of the correlation coefficients of any two data in the current crystallization stage and the next crystallization stage is calculated. The differential vector is weighted by using the normalized value of the absolute value of the ratio to obtain a predicted differential vector of the next crystallization stage; The column vector corresponding to the next crystallization stage in the process memory matrix is ​​used as a reference vector, and based on the distance between the differential vector of the next crystallization stage and the same data in the reference vector, the similarity between the two corresponding vectors is measured, and the similarity is negatively correlated with the distance; Calculate the degree of abnormality in the next crystallization stage to satisfy the following relationship: ; In the formula, For the The degree of abnormality of each crystallization stage; For the The similarity between the difference vector of each crystallization stage and the reference vector; For the The difference vector of the crystallization stage; For the The predicted difference vector of each crystallization stage; is the representation of cosine similarity; is the natural exponential function; is an absolute value symbol; based on the comparison result between the abnormality degree and the preset threshold value, the state of the crystallizer is monitored.

2. A method for monitoring the operation status of a crystallizer according to claim 1, characterized in that: The normalized value is used as the data change synchronization rate between the current crystallization stage and the next crystallization stage. The data change synchronization rate satisfies the following relationship: ； In the formula, For the The crystallization stage and The data change synchronization rate of each crystallization stage; , Respectively The crystallization stage and Correlation coefficient between temperature data and pressure data in each crystallization stage; , Respectively The crystallization stage and Correlation coefficient between temperature data and vibration data in each crystallization stage; , Respectively The crystallization stage and Correlation coefficient between pressure data and vibration data in each crystallization stage; is the absolute value symbol; is the normalization function.

3. A method for monitoring the operation status of a crystallizer according to claim 2, characterized in that: The correlation coefficient is a Pearson correlation coefficient or a Spearman rank correlation coefficient.

4. A method for monitoring the operation status of a crystallizer according to claim 1, characterized in that: The similarity degree satisfies the following relationship: ； In the formula, For the The similarity between the difference vector of each crystallization stage and the reference vector; For the The difference vector of the crystallization stage elements; For the The reference vector of the crystallization stage elements; for the summation symbol; is the normalization function.

5. A method for monitoring the operation status of a crystallizer according to claim 1, characterized in that: The monitoring of the state of the crystallizer based on the comparison result of the abnormality degree and the preset threshold value includes: Obtaining a preset threshold; if the degree of abnormality in the next crystallization stage of the current crystallization stage is greater than or equal to the preset threshold, determining that the operation state of the crystallizer in the next crystallization stage is abnormal; If it is less than the preset threshold, it is determined that the operation state of the crystallizer in the next crystallization stage is normal.

6. A crystallizer operation status monitoring system, characterized in that: The crystallizer operation status monitoring system includes a memory and a processor, the memory stores a computer program, and the processor executes the computer program to implement the steps of the crystallizer operation status monitoring method as described in any one of claims 1 to 5.

Images