Self-adaptive and self-decision-making traditional Chinese medicine extraction process boiling state judgment method, storage medium and electronic device

Abstract

The invention discloses a self-adaptive and self-decision-making traditional Chinese medicine extraction process boiling state judgment method, a storage medium and an electronic device, and belongs to the technical field of traditional Chinese medicine extracting.According to the method, an acoustic emission technology is used for monitoring the boiling behavior in the traditional Chinese medicine extraction process, and the corresponding relation between an AE signal and the boiling state is established; the invention further provides a self-adaptive self-decision-making method based on iterative multivariable statistical analysis, and the method is used for on-line monitoring and boiling state judgment of boiling states in different medicinal material extraction systems. The method is high in applicability, priori database reconstruction does not need to be carried out for different medicinal material extraction systems, maintenance work of later model correction and updating is avoided, self-adaptive self-decision online judgment of boiling states in different extraction systems can be achieved, the invalid heating period is shortened, production energy consumption is reduced, and the method is suitable for large-scale popularization and application. And a key technical support is provided for intelligent manufacturing transformation of traditional Chinese medicines.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of traditional Chinese medicine extraction, and particularly relates to a self-adaptive self-decision traditional Chinese medicine extraction process boiling state judgment method, a storage medium and an electronic device. BACKGROUND

[0002] Extraction is one of the core links of traditional Chinese medicine production process, and the process control level directly affects the quality and energy efficiency level of the final product. However, the boiling state recognition of the extraction process still has problems of lag in boiling time judgment and high dependence on manual experience at the present stage, lacks online monitoring technology, and the intelligent level needs to be improved. Specifically, traditional Chinese medicine pharmaceutical enterprises generally use two-stage extraction process, that is, first heating with "strong fire", adjusting the steam valve to reduce "fire" when boiling, and then maintaining "micro-boiling" with "gentle fire" until the process specified time. The so-called "micro-boiling" state actually means that the traditional Chinese medicine extraction system has entered saturated boiling, not subcooled boiling. The boiling state detection in industry mainly relies on two traditional judgment methods: one is to detect whether the liquid reaches the saturated temperature through the temperature sensor built-in the extraction tank, and the other is for the operator to observe the bubble shape through the viewing window for experience judgment. However, the traditional method has obvious technical limitations. When the worker judges the boiling state by experience, he is often affected by the light conditions in the tank, the shielding of suspended materials and the interference of floating scum, resulting in subjective and uncertain judgment results. Since the industrial production tank is large and the temperature is not uniform everywhere, when the temperature sensor detects that the system temperature reaches saturated boiling, the system has actually been boiling for some time. The lag in boiling state determination will cause unnecessary "strong fire" heating, increase steam consumption, and prolong the production cycle of the extraction process. Therefore, in order to improve production efficiency, reduce energy consumption, and realize precise control of the process, it is urgent to develop an online analysis method for accurately detecting the boiling state to realize intelligent upgrading of the extraction process.

[0003] Acoustic emission (AE) technology is a non-invasive process monitoring method that can sense transient elastic waves generated when energy is rapidly released in a dynamic process. In recent years, AE technology has been widely used in state monitoring due to its convenient installation and high sensitivity. For example, the Chinese patent document with publication number CN118670863A discloses a micro-motion fatigue state monitoring method based on acoustic emission technology, and the Chinese patent document with publication number CN109813805A discloses a laser cleaning process monitoring method based on acoustic emission technology.

[0004] The nucleation, growth, detachment, oscillation and collapse of the bubbles in each stage can stimulate acoustic characteristics in specific frequency bands, and the bubble behavior dynamics correspond to different boiling states, so the spectral characteristics of the AE signals can be used to identify various states in the boiling system. A Chinese patent document with publication number CN103115936A discloses a boiling state detection method, which selects the total energy, standard deviation, average absolute deviation and main frequency of the acoustic signals as characteristic parameters, analyzes the change characteristics of the characteristic parameters, and uses the main frequency value change related parameter k value as a criterion to judge the boiling state and its transition, but this method can only be used for water boiling state detection. The extraction of traditional Chinese medicine is more complex, not only a solid-liquid mixed system, but also the roughness of the medicinal material surface, the geometric shape, the chemical component content in the extraction liquid (such as saponin components which may act as a surfactant), the extraction solvent (such as the proportion of ethanol) and other factors will affect the AE signal spectral characteristics, and the method in the above patent cannot be used for boiling state detection in different medicinal material extraction systems. Therefore, it is necessary to develop an accurate boiling state judgment method for traditional Chinese medicine extraction process with universality. SUMMARY

[0005] In order to solve the problems existing in the prior art, the present application provides a self-adaptive self-decision traditional Chinese medicine extraction process boiling state judgment method, which has strong applicability, does not need to reconstruct prior database for different medicinal material extraction systems, avoids the maintenance work of later model correction and update, and can realize self-adaptive self-decision online judgment of boiling state in different extraction systems, shorten the invalid heating period, reduce the production energy consumption, and provide key technical support for traditional Chinese medicine intelligent manufacturing transformation. The present application adopts acoustic emission technology, the sensor can be attached to the outside of the extraction tank to detect signals, without modifying the existing industrial extraction tank, and can realize real-time nondestructive determination, and has good popularization prospect.

[0006] The specific technical solutions adopted are as follows:

[0007] A self-adaptive self-decision traditional Chinese medicine extraction process boiling state judgment method, comprising the following steps:

[0008] Step 1: Use the acoustic emission signal acquisition system to monitor the acoustic emission signal change in the traditional Chinese medicine extraction process, collect the acoustic emission signal in a fixed time period, perform spectral analysis on the acoustic emission signal obtained each time to obtain an acoustic emission spectrum, obtain N acoustic emission spectra and number them in time sequence, and when N≥10, perform steps 2-6;

[0009] Step 2: Perform principal component analysis on the first to N-1 acoustic emission spectra obtained in step 1 to establish a first PCA model and determine the batch control limit T 2 limit N-1 ;

[0010] Step 3: Project the Nth acoustic emission spectrum into the first PCA model in real time, and calculate its Hotelling's T 2 statistic T 2 N-1 , and compare it with T 2 N-1 . 2 limit N-1 .

[0011] Step 4: Perform principal component analysis on the 1st to Nth acoustic emission spectra obtained in Step 1, establish a second PCA model, and determine the batch control limit T 2 limit N .

[0012] Step 5: Project the Nth acoustic emission spectrum into the second PCA model in real time, and calculate its Hotelling's T 2 statistic T 2 N , and compare it with T 2 N . 2 limit N .

[0013] Step 6: If T 2 N-1 ≤ T 2 limit N-1 or T 2 N ≤ T 2 limit N , continue to collect acoustic emission signals according to the method of Step 1, obtain acoustic emission spectra and number them in time sequence, update the value of N, and repeat Steps 2 to 6; when T 2 N-1 > T 2 limit N-1 and T 2 N > T 2 limit N , it is judged to be the boiling state.

[0014] The method of the present application uses acoustic emission (AE) technology to monitor the boiling behavior in the extraction process of traditional Chinese medicine. By analyzing the AE signal and the change of boiling bubble behavior during the boiling process, the corresponding relationship between AE signal and boiling state is established, and an adaptive self-decision method based on iterative multivariate statistical analysis is proposed for online monitoring and boiling state judgment in different medicinal material extraction systems.

[0015] Optionally, in step 1, the acoustic emission signal is collected at the beginning of the traditional Chinese medicine extraction process.

[0016] Further, the acoustic emission signal collection system is composed of an acoustic emission sensor, an acoustic emission signal instrument and a data acquisition card.

[0017] Further, during the collection of the acoustic emission signal, the acoustic emission signal instrument is set to have an amplification factor of 1-10000, a sampling rate of 0.2-1MHz, a signal collection period of 30-60s and a collection duration of 5-20s.

[0018] Preferably, the acoustic emission signal instrument is set to have an amplification factor of 1000, a sampling rate of 1MHz, a signal collection period of 30s and a collection duration of 10s.

[0019] Specifically, the traditional Chinese medicine extraction process in step 1 includes a single medicinal material extraction process or a compound medicinal material extraction process. The method of the present application has a wide range of applications and can be used for boiling state judgment in different traditional Chinese medicine extraction systems.

[0020] Specifically, in steps 2 and 4, the power spectral density (sound signal intensity at a specific frequency) in the frequency band of 75-100kHz in each acoustic emission spectrum is selected to establish the first PCA model and the second PCA model, respectively, and the dimensionality is reduced by principal component analysis, and the principal components with an explanation variation ability greater than 0.05 are retained for calculating the Hotelling’s T 2 statistic and determining the batch control limit.

[0021] Specifically, in steps 2 and 4, the Hotelling’s T 2 statistic is calculated according to the retained principal components, and the Hotelling’s T 2 statistic is the common accumulation of the normalized scores of all retained principal components, and the Hotelling’s T 2 statistic of the nth sampling point is calculated according to the formula (1), and the Hotelling’s T 2 statistic is used to determine whether the observation value has a significant deviation from the center position of other observation values, wherein A represents the number of retained principal components.

[0022] ​

[0023] In formula (1), t n is a vector composed of A principal component scores of the nth sampling point (the total number of sampling points is N), and is a vector composed of the mean value of the score of each principal component at all sampling points, λ is a diagonal matrix composed of the eigenvalues corresponding to the A principal components, and the upper right T superscript represents the transpose operation;

[0024] Hotelling's T 2 Batch control limit of the statistical quantity The F distribution is calculated by formula (2):

[0025]

[0026] A represents the number of retained principal components, N represents the total number of sampling points, F 1-α (A, N-A) is the upper critical value of the F distribution with degrees of freedom (A, N-A) at the significance level α.

[0027] Further, the boiling state is divided into two stages of subcooled boiling and saturated boiling, when T 2 N-1 T 2 limit N-1 T 2 N T 2 limit N When T 2 N-1 T 2 limit N-1 T 2 N T 2 limit N When T

[0028] The application further provides a storage medium, wherein the storage medium stores a program, and the program performs the adaptive self-decision Chinese medicine extraction process boiling state judgment method when running.

[0029] The application further provides an electronic device, which comprises a memory, a processor, and a computer program stored in the memory and capable of running on the processor, and the processor performs the adaptive self-decision Chinese medicine extraction process boiling state judgment method through the computer program.

[0030] Compared with the prior art, the application has the following beneficial effects:

[0031] (1) In the existing boiling state judgment method, the temperature monitoring method is limited by thermal inertia and spatial heterogeneity, and the monitoring results often have hysteresis, and the boiling temperature of different extraction systems is different, so it is impossible to develop a unified temperature standard for boiling state judgment. The naked eye observation method is easily affected by the light in the tank, the blocking of the medicinal material foam and the like, and is highly subjective, and the judgment result is difficult to standardize. Compared with the above, the boiling state judgment based on the AE signal of the present application can directly perceive the phase change process, overcome the lag of heat conduction, provide more accurate and stable boiling state criterion for the industrialized production of traditional Chinese medicine, shorten the invalid heating period, and reduce the production energy consumption. In addition, the acoustic emission sensor can be attached to the outside of the extraction tank to detect the signal, without modifying the existing industrial extraction tank, and can realize real-time nondestructive determination, and has good popularization prospect.

[0032] (2) The present application adopts Hotelling's T 2 The Hotelling's T2statistic monitors the power spectral density at multiple frequencies in the frequency band, and updates the control limit in real time through an iterative modeling method to eliminate the influence of batch differences on the model. This adaptive self-decision method does not need to construct a priori database, avoiding the maintenance work of model correction and update in the later stage. BRIEF DESCRIPTION OF DRAWINGS

[0033] Figure 1 Figure 1 is a flowchart of the boiling state judgment method for the self-adaptive self-decision traditional Chinese medicine extraction process in the present application.

[0034] Figure 2 Figure 2 is a graph showing the changes of temperature and AE signal at different boiling stages in the extraction process of different single medicinal materials, wherein (a) is Radix Rehmanniae Preparata, (b) is Phellodendri Ammonium, (c) is Anemarrhena Asphodeloides, and (d) is Prunella Vulgaris.

[0035] Figure 3 Figure 3 is a boiling judgment result graph of the method of the present application in the extraction process of different medicinal materials, wherein (a) is Radix Rehmanniae Preparata, (b) is Phellodendri Ammonium, (c) is Anemarrhena Asphodeloides, and (d) is Prunella Vulgaris.

[0036] Figure 4 Figure 4 is a graph showing the changes of temperature and AE signal at different boiling stages in the extraction process of a compound.

[0037] Figure 5 Figure 5 is a boiling judgment result graph of the method of the present application in the extraction process of a compound. DETAILED DESCRIPTION

[0038] The present application will be further illustrated in conjunction with the following examples. It should be understood that these examples are only used to illustrate the present application, and are not used to limit the scope of the present application.

[0039] Example 1

[0040] The flowchart of the self-adaptive self-decision traditional Chinese medicine extraction process boiling state judgment method in the application is shown in Figure Figure 1 , which comprises the following steps:

[0041] Step 1: The acoustic emission signal acquisition system composed of an acoustic emission sensor, an acoustic emission signal instrument and a data acquisition card is used to acquire the acoustic emission signal in the traditional Chinese medicine extraction process. During the acquisition of the acoustic emission signal, the acoustic emission signal instrument is set to have an amplification of 1000, a sampling rate of 1 MHz and a signal acquisition period of 30 s, and the acquisition time is 10 s. The frequency spectrum of the acoustic emission signal obtained by each acquisition is analyzed, specifically, the time-domain signal obtained by the acquisition is analyzed by discrete Fourier transform, and the calculation process includes windowing, Fourier transform, power spectrum density estimation and normalization processing, so as to obtain the energy distribution of the signal at different frequencies, obtain the acoustic emission frequency spectrum, and number the acoustic emission frequency spectrum according to the time sequence number 1-N. When N≥10, steps 2-6 are performed.

[0042] Step 2: The first to N-1 acoustic emission frequency spectra obtained in step 1 are analyzed by principal component analysis, a first PCA model is established, and a batch control limit T 2 limit N-1 is determined.

[0043] Step 3: The N-1 acoustic emission frequency spectrum is projected into the first PCA model in real time, and its Hotelling’s T 2 statistic T 2 N-1 is calculated. T 2 N-1 is compared with T 2 limit N-1 .

[0044] Step 4: The first to N acoustic emission frequency spectra obtained in step 1 are analyzed by principal component analysis, a second PCA model is established, and a batch control limit T 2 limit N is determined.

[0045] Step 5: The N acoustic emission frequency spectrum is projected into the second PCA model in real time, and its Hotelling’s T 2 statistic T 2 N is calculated. T 2 N is compared with T 2 limit N .

[0046] Step 6: if T 2 N-1 ≤ T 2 limit N-1 or T 2 N ≤ T 2 limit N , then continue to collect the acoustic emission signals according to the method of Step 1, obtain the acoustic emission spectrum and number in time sequence, update the value of N, and repeat Steps 2-6; when T 2 N-1 > T 2 limit N-1 and T 2 N > T 2 limit N , it is judged as boiling state.

[0047] Further, in Steps 2 and 4, the power spectral density in the frequency band of 75-100 kHz in each acoustic emission spectrum is selected to establish the first PCA model and the second PCA model, respectively, dimension reduction is performed through principal component analysis, and the principal components with the ability to explain the variation greater than 0.05 are retained for calculating the Hotelling’s T 2 statistic and determining the batch control limit T 2 limit .

[0048] In Steps 2 and 4, the Hotelling’s T 2 statistic is calculated according to the retained principal components, the Hotelling’s T 2 statistic is the sum of the normalized scores of all the retained principal components, and the Hotelling’s T 2 statistic of the nth sampling point is calculated according to the formula (1), and the Hotelling’s T 2 statistic is used to determine whether the observation value deviates significantly from the center position of other observation values, wherein A represents the number of retained principal components.

[0049]

[0050] In formula (1), t n is a vector composed of A principal component scores of the nth sampling point, is a vector composed of the mean value of the scores of each principal component at all sampling points, and λ is a diagonal matrix composed of the characteristic values corresponding to the A principal components.

[0051] Hotelling’s T​2 Batch control limits for statistics The F-distribution of formula (2) is used to calculate:

[0052]

[0053] A represents the number of principal components retained, N represents the total number of sampling points, and F 1-α (A, NA) is the upper critical value of the significance level α for an F distribution with (A, NA) degrees of freedom.

[0054] Example 2

[0055] The method in Example 1 was applied to the extraction process of single-herb medicinal materials.

[0056] Based on the observed bubble state, the extraction process was divided into two stages: supercooled boiling and saturated boiling. Specifically, the observation that bubbles could rise to the surface of the liquid in the extraction tank and burst was used as the indicator of saturated boiling. It was found that the Rehmannia glutinosa extraction system reached saturated boiling at 19.0 min (92.0℃), the Phellodendron chinense extract system reached saturated boiling at 18.5 min (91.0℃), the Anemarrhena asphodeloides extract system reached saturated boiling at 18.5 min (90.5℃), and the Prunella vulgaris extract system reached saturated boiling at 15.5 min (85.4℃). Figure 2 The results in (a)-(d) show that there is a correspondence between the changes in acoustic emission signals and different boiling states during the extraction of single medicinal materials, and that the changes in acoustic emission signals can be used to determine the boiling state.

[0057] The boiling state was monitored in real time using the method described in Example 1, and the boiling determination results are as follows: Figure 3 As shown in (a)-(d) in the figure, during the water extraction process of different medicinal materials, when two consecutive time points Hotelling's T 2 The absolute error between the time when the statistic exceeds the control limit and the time when the system reaches saturated boiling is less than 2 minutes. Specifically, the Rehmannia glutinosa extract system was determined to be boiling at 20.5 minutes and reached saturated boiling at 19.0 minutes; the Phellodendron amurense extract system was determined to be boiling at 20.5 minutes and reached saturated boiling at 18.5 minutes; the Anemarrhena asphodeloides extract system was determined to be boiling at 17.5 minutes and reached saturated boiling at 18.5 minutes; and the Prunella vulgaris extract system was determined to be boiling at 17.0 minutes and reached saturated boiling at 15.5 minutes. These results indicate that the prediction results of this method are relatively accurate. Furthermore, this method does not require re-screening for optimal frequencies for different herbal extract systems, nor does it require establishing prior batch control limits. It can achieve online boiling determination in relatively complex herbal-liquid systems, demonstrating simplicity and good applicability.

[0058] Example 3

[0059] The method in Example 1 was applied to the extraction process of compound medicinal materials (Radix Rehmanniae Preparata, Salt Anemarrhena Rhizome, Salt Cinnamon).

[0060] In order to investigate the effectiveness and applicability of the method in compound extraction system, it was applied to monitor the boiling judgment of the extraction process of Radix Rehmanniae Preparata, Salt Anemarrhena Rhizome and Salt Cinnamon. Figure 4 The results in Example 3 show that there is a corresponding relationship between the change of acoustic emission signal and different boiling states in the extraction process of compound medicinal materials, and the change of acoustic emission signal can be used for boiling state judgment.

[0061] The method in Example 1 was used for real-time monitoring of boiling state, and the boiling judgment result is shown in Table 2. Figure 5 As shown in Table 2, the method judges boiling at 19.5 min, indicating that the system has entered saturated boiling at this moment, and the heating medium temperature can be reduced to maintain the extraction state of "gentle simmer". Compared with the traditional temperature detection standard, this method can identify that the system has entered the saturated boiling stage 8.5 min earlier, shortens the heating stage time by 30% in the extraction process, and improves the production efficiency. The shortening of the heating stage time also reduces the heating in the extraction process, which is conducive to energy saving.

[0062] The above examples have described the technical solutions of the present application in detail. It should be understood that the above description is only a specific embodiment of the present application and is not used to limit the present application. Any modification, supplement or similar replacement within the principle range of the present application should be included in the protection scope of the present application.

Claims

1. An adaptive and self-decision-making method for determining the boiling state in the extraction process of traditional Chinese medicine, characterized in that, Includes the following steps: Step 1: Use an acoustic emission signal acquisition system to monitor the changes in acoustic emission signals during the extraction of traditional Chinese medicine. Acquire acoustic emission signals at fixed time periods. Perform spectrum analysis on each acquired acoustic emission signal to obtain an acoustic emission spectrum. Obtain N acoustic emission spectra and number them according to the time sequence. When N≥10, proceed to steps 2-6. Step 2: Perform principal component analysis on the acoustic emission spectra from the 1st to the (N-1th)th obtained in Step 1, establish the first PCA model, and determine the batch control limit T. 2 limit N-1 ; Step 3: Project the (N-1)th acoustic emission spectrum into the first PCA model in real time, and calculate its Hotelling's T. 2 Statistic T 2 N-1 , will T 2 N-1 With T 2 limit N-1 Compare; Step 4: Perform principal component analysis on the acoustic emission spectra obtained in Step 1 from the 1st to the Nth, establish a second PCA model, and determine the batch control limit T. 2 limit N ; Step 5: Project the Nth acoustic emission spectrum into the second PCA model in real time, and calculate its Hotelling's T. 2 Statistic T 2 N , will T 2 N With T 2 limit N Compare; Step 6: If T 2 N-1 ≤T 2 limit N-1 or T 2 N ≤T 2 limit N Then, continue collecting acoustic emission signals according to step 1, obtain the acoustic emission spectrum, number it according to the time sequence, update the N value, and repeat steps 2 to 6; when T 2 N-1 >T 2 limit N-1 And T 2 N >T 2 limit N If it is, then it is judged to be in a boiling state.

2. The adaptive self-decision-making method for judging the boiling state in the extraction process of traditional Chinese medicine according to claim 1, characterized in that, The acoustic emission signal acquisition system consists of an acoustic emission sensor, an acoustic emission modulation instrument, and a data acquisition card.

3. The adaptive self-decision-making method for judging the boiling state in the extraction process of traditional Chinese medicine according to claim 2, characterized in that, During the acquisition of acoustic emission signals, the amplification factor of the acoustic emission signal modulator was set to 1-10000, and the sampling rate was 0.2-1MHz.

4. The adaptive self-decision-making method for judging the boiling state in the extraction process of traditional Chinese medicine according to claim 1, characterized in that, In steps 2 and 4, the power spectral density within the 75-100kHz frequency band of each acoustic emission spectrum is selected to establish the first PCA model and the second PCA model, respectively. Principal component analysis is used for dimensionality reduction, and principal components with an explanatory power greater than 0.05 are retained for calculating Hotelling's T. 2 Statistics and determination of batch control limits.

5. The adaptive self-decision-making method for judging the boiling state in the extraction process of traditional Chinese medicine according to claim 4, characterized in that, In steps 2 and 4, Hotelling's T is calculated based on the retained principal components. 2 Statistics, Hotelling's T 2 The statistic is obtained by summing the normalized scores of all retained principal components, where Hotelling'sT is the sum of the scores of the nth sampling point. 2 Statistic The calculation formula is shown in formula (1), Hotelling's T 2 The statistic is used to determine whether an observation deviates significantly from the central location of other observations, where A represents the number of principal components retained; In formula (1), t n Let A be the vector formed by the scores of the principal components at the nth sampling point. λ is a vector composed of the average scores of each principal component across all sampling points, and λ is a diagonal matrix composed of the eigenvalues ​​corresponding to the A principal components. Hotelling's T 2 Batch control limits for statistics The F-distribution of formula (2) is used to calculate: A represents the number of principal components retained, N represents the total number of sampling points, and F 1-α (A, NA) is the upper critical value of the significance level α for an F distribution with (A, NA) degrees of freedom.

6. The adaptive self-decision-making method for judging the boiling state in the extraction process of traditional Chinese medicine according to claim 1, characterized in that, Boiling states are divided into two states: supercooled boiling and saturated boiling. When T 2 N-1 >T 2 limit N-1 And T 2 N >T 2 limit N The boiling state determined at that time is saturated boiling.

7. A storage medium, characterized in that, The storage medium contains a program, wherein when the program runs, it executes the adaptive self-decision-making method for judging the boiling state of the traditional Chinese medicine extraction process as described in any one of claims 1-6.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, The processor executes the adaptive self-decision-making method for judging the boiling state of the traditional Chinese medicine extraction process according to any one of claims 1-6 through the computer program.

Images