Adaptive self-decision traditional chinese medicine extraction process boiling state judgment method, storage medium and electronic device

Abstract

The application discloses a self-adaptive self-decision traditional Chinese medicine extraction process boiling state judgment method, a storage medium and an electronic device, belongs to the technical field of traditional Chinese medicine extraction, and the method uses acoustic emission technology to monitor the boiling behavior in the traditional Chinese medicine extraction process, establishes the corresponding relationship between AE signals and boiling states, and further proposes a self-adaptive self-decision method based on an iterative multivariate statistical analysis, which is used for online monitoring and boiling state judgment of boiling states in different medicine extraction systems. The method has strong applicability, does not need to reconstruct a prior database for different medicine extraction systems, avoids maintenance work of model correction and update in the later period, and can realize self-adaptive self-decision online judgment of boiling states in different extraction systems, shortens the invalid heating period, reduces production energy consumption, and provides key technical support for intelligent manufacturing transformation of traditional Chinese medicines.

Description

Technical Field

[0001] This invention belongs to the field of traditional Chinese medicine extraction technology, specifically relating to an adaptive and self-decision-making method for judging the boiling state in the extraction process of traditional Chinese medicine, a storage medium, and an electronic device. Background Technology

[0002] Extraction is one of the core processes in the production of traditional Chinese medicine (TCM), and its process control level directly affects the quality and efficiency of the final product. However, at present, the identification of the boiling state in the extraction process still suffers from problems such as delayed judgment of the boiling point and heavy reliance on human experience, lack of online monitoring technology, and insufficient level of intelligence. Specifically, TCM pharmaceutical companies generally adopt a two-stage extraction process: first, heating with a "high heat" to reduce the "heat" by adjusting the steam valve when boiling, and then maintaining a "gentle boil" until the specified time. The so-called "gentle boil" state actually refers to the TCM extraction system reaching saturated boiling, rather than supercooled boiling. In industry, the detection of the boiling state mainly relies on two traditional methods: one is to detect whether the liquid has reached the saturation temperature through a temperature sensor built into the extraction tank, and the other is for operators to make an experience-based judgment by observing the bubble shape through a viewing window. However, traditional methods have obvious technical limitations. When workers judge the boiling state based on experience, they are often affected by the lighting conditions inside the tank, the obstruction of suspended matter in the medicinal materials, and the interference of foam, resulting in subjective and uncertain judgments. Because industrial production tanks are large and temperatures are uneven, by the time a temperature sensor detects that the system has reached saturated boiling, it has often already been boiling for some time. This lag in boiling state determination leads to unnecessary high-intensity heating, increasing steam consumption and prolonging the extraction process cycle. Therefore, to improve production efficiency, reduce energy consumption, and achieve precise process control, it is urgent to develop an accurate online analysis method for detecting boiling state, thereby enabling intelligent upgrades to 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 during dynamic processes. In recent years, AE technology has been widely used in condition monitoring due to its advantages such as convenient installation and high sensitivity. For example, Chinese patent document CN118670863A discloses a micro-motion fatigue condition monitoring method based on acoustic emission technology, and Chinese patent document CN109813805A discloses a laser cleaning process monitoring method based on acoustic emission technology.

[0004] The nucleation, growth, detachment, oscillation, and collapse of bubbles excite specific frequency bands of acoustic features. Since bubble dynamics correspond to different boiling states, the spectral characteristics of the acoustic emission (AE) signal can be used to identify various states in a boiling system. For example, Chinese patent document CN103115936A discloses a method for detecting boiling states. This method selects the total energy, standard deviation, mean absolute deviation, and dominant frequency of the acoustic signal as characteristic parameters, analyzes the changes in these parameters, and uses the parameter k, which is related to the change in the dominant frequency, as a criterion to determine the boiling state and its transitions. However, this method can only be used to detect the boiling state of water. The situation in traditional Chinese medicine extraction is more complex. It involves not only solid-liquid mixtures but also factors such as the surface roughness and geometry of the medicinal materials, the content of chemical components in the extract (e.g., saponins may act as surfactants), and the extraction solvent (e.g., the proportion of ethanol). These factors all affect the spectral characteristics of the AE signal. The method in the aforementioned patent cannot be used to detect the boiling state in different medicinal material extraction systems. Therefore, it is necessary to develop a universally applicable and accurate method for determining the boiling state in the traditional Chinese medicine extraction process. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides an adaptive, self-determining method for determining the boiling state in the extraction process of traditional Chinese medicine. This method is highly applicable, eliminates the need for prior database reconstruction for different medicinal material extraction systems, and avoids the maintenance work of subsequent model calibration and updates. Furthermore, this method can achieve adaptive, self-determining online determination of the boiling state in different extraction systems, shortening the ineffective heating cycle, reducing production energy consumption, and providing key technical support for the intelligent manufacturing transformation of traditional Chinese medicine. This invention employs acoustic emission technology; the sensor can be attached to the outside of the extraction tank to detect signals without modifying existing industrial extraction tanks, enabling real-time, non-destructive measurement and showing good prospects for widespread application.

[0006] The specific technical solution adopted is as follows:

[0007] An adaptive and self-decision-making method for determining the boiling state in the extraction process of traditional Chinese medicine includes the following steps:

[0008] 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.

[0009] 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 ;

[0010] 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;

[0011] 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 ;

[0012] Step 5: Project the Nth acoustic emission spectrum into the second PCA model in real time to calculate its Hotelling's T. 2 Statistic T 2 N , will T 2 N With T 2 limit N Compare;

[0013] 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 the temperature is high, it is judged to be in a boiling state.

[0014] The present invention uses acoustic emission (AE) technology to monitor boiling behavior during the extraction process of traditional Chinese medicine. By analyzing the AE signal and the changes in boiling bubble behavior during the boiling process, a correspondence between the AE signal and the boiling state is established. Furthermore, an adaptive self-decision-making 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, acoustic emission signals are collected at the beginning of the traditional Chinese medicine extraction process.

[0016] Furthermore, the acoustic emission signal acquisition system consists of an acoustic emission sensor, an acoustic emission modulation instrument, and a data acquisition card.

[0017] Furthermore, during the acquisition of acoustic emission signals, the amplification factor of the acoustic emission signal modulator is set to 1-10000, the sampling rate is 0.2-1MHz, the signal acquisition period is 30-60s, and the acquisition duration is 5-20s.

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

[0019] Specifically, the traditional Chinese medicine extraction process in step 1 includes extraction of single-herb ingredients or extraction of compound herbs. The method of this invention has a wide range of applications and can be used to determine the boiling state of different traditional Chinese medicine extraction systems.

[0020] Specifically, in steps 2 and 4, the power spectral density (sound signal intensity at a specific frequency) 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 to calculate Hotelling's T. 2 Statistics and determination of batch control limits.

[0021] Specifically, 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, and the Hotelling's T for the nth sampling point is... 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;

[0022]

[0023] In formula (1), t n Let A be the vector formed by the scores of the principal components of the nth sampling point (the total number of sampling points is N). It is a vector composed of the average scores of each principal component across all sampling points, λ is a diagonal matrix composed of the eigenvalues ​​corresponding to the A principal components, and the superscript T in the upper right corner indicates the transpose operation;

[0024] Hotelling's T 2 Batch control limits for statistics The F-distribution of formula (2) is used to calculate:

[0025]

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

[0027] Furthermore, the boiling state is divided into two stages: 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 by time T is saturated boiling. In the extraction process of traditional Chinese medicine, when T... 2 N-1 >T 2 limit N-1 And T 2 N >T 2 limit N At this time, the heat during the extraction process of traditional Chinese medicine can be reduced to maintain a gentle boil for the specified time.

[0028] The present invention also provides a storage medium storing a program, wherein the program executes the adaptive self-decision-making method for judging the boiling state of the traditional Chinese medicine extraction process when it runs.

[0029] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the adaptive self-decision-making method for judging the boiling state of the traditional Chinese medicine extraction process through the computer program.

[0030] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0031] (1) Among existing methods for determining boiling state, temperature monitoring is limited by thermal inertia and spatial heterogeneity, resulting in lag in monitoring results. Furthermore, the boiling temperatures of different extraction systems vary, making it impossible to establish a unified temperature standard for boiling state determination. Visual observation is susceptible to interference from light within the tank and obstruction by herbal foam, and is highly subjective, making standardization difficult. In contrast, this invention uses acoustic emission (AE) signals for boiling state determination, directly sensing the phase transition process and overcoming thermal conduction lag. This provides a more accurate and robust criterion for boiling state determination in the industrial production of traditional Chinese medicine, shortening the ineffective heating cycle and reducing production energy consumption. Moreover, the acoustic emission sensor can be attached to the outside of the extraction tank to detect signals, eliminating the need to modify existing industrial extraction tanks and enabling real-time, non-destructive measurement, thus showing promising prospects for widespread application.

[0032] (2) This invention adopts Hotelling's T 2 The statistics simultaneously monitor the power spectral density at multiple frequencies within the frequency band, and update the control limits in real time through iterative modeling to eliminate the impact of batch differences on the model. This adaptive self-decision-making method eliminates the need to build a prior database, avoiding the maintenance work of later model correction and updates. Attached Figure Description

[0033] Figure 1 This is a flowchart illustrating the adaptive self-decision-making method for determining the boiling state in the extraction process of traditional Chinese medicine in this invention.

[0034] Figure 2 The graph shows the changes in temperature and AE signal at different boiling stages during the extraction process of different single-herb medicinal materials. (a) is Rehmannia glutinosa (processed), (b) is Phellodendron chinense (processed with salt), (c) is Anemarrhena asphodeloides (processed with salt), and (d) is Prunella vulgaris (processed with salt).

[0035] Figure 3 The diagram shows the boiling results of the extraction process of different medicinal materials using the method of the present invention. (a) is Rehmannia glutinosa (processed), (b) is Phellodendron chinense (processed with salt), (c) is Anemarrhena asphodeloides (processed with salt), and (d) is Prunella vulgaris (processed with salt).

[0036] Figure 4 The graph shows the changes in temperature and AE signal at different boiling stages during the compound extraction process.

[0037] Figure 5 This is a diagram showing the boiling determination results of the method of the present invention during the compound extraction process. Detailed Implementation

[0038] The present invention will be further illustrated below with reference to the embodiments. It should be understood that these embodiments are for illustrative purposes only and are not intended to limit the scope of the invention.

[0039] Example 1

[0040] The flowchart of the adaptive self-decision-making method for determining the boiling state in the extraction process of traditional Chinese medicine in this invention is shown below. Figure 1 As shown, it includes the following steps:

[0041] Step 1: Acoustic emission signal acquisition system consisting of acoustic emission sensor, acoustic emission signal modulator and data acquisition card is used to collect acoustic emission signals during the extraction process of traditional Chinese medicine. During the acquisition of acoustic emission signals, the amplification factor of acoustic emission signal modulator is set to 1000, the sampling rate is 1MHz, the signal acquisition period is 30s, and the acquisition time is 10s. Spectral analysis is performed on each acquired acoustic emission signal to obtain an acoustic emission spectrum. Specifically, in the acoustic emission signal processing, the acquired time-domain signal is subjected to spectral analysis through discrete Fourier transform. The calculation process includes windowing, Fourier transform, power spectral density estimation and normalization to obtain the energy distribution of the signal at different frequencies and obtain the acoustic emission spectrum. The acoustic emission spectrum is numbered from 1 to N according to the time sequence number. When N≥10, proceed to steps 2-6.

[0042] 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 ;

[0043] 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;

[0044] 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 ;

[0045] Step 5: Project the Nth acoustic emission spectrum into the second PCA model in real time to calculate its Hotelling's T. 2 Statistic T 2 N , will T 2 N With T 2 limit N Compare;

[0046] 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 the temperature is high, it is judged to be in a boiling state.

[0047] Furthermore, in steps 2 and 4, the power spectral density within the 75-100kHz frequency band of each acoustic emission spectrum is selected to establish a first PCA model and a 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 T 2 limit ;

[0048] In steps 2 and 4, Hotelling's T is calculated based on the retained principal components. 2 Statistics, Hotelling's 2 The statistic is obtained by summing the normalized scores of all retained principal components, and the Hotelling's T for the nth sampling point is... 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;

[0049]

[0050] 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.

[0051] Hotelling's T2 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 represents the number of principal components retained. 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 (Rehmannia glutinosa, Anemarrhena asphodeloides, and Phellodendron chinense).

[0060] To examine the effectiveness and applicability of the method in compound extraction systems, it was applied to monitor the boiling point of the compound extraction process of Rehmannia glutinosa, Anemarrhena asphodeloides (processed with salt), and Phellodendron chinense (processed with salt). Figure 4 The results show that there is a correspondence between changes in acoustic emission signals and different boiling states during the extraction of compound medicinal materials, and that changes in acoustic emission signals can be used to determine the boiling state.

[0061] The boiling state was monitored in real time using the method described in Example 1, and the boiling determination results are as follows: Figure 5 As shown, this method determines boiling at 19.5 minutes, indicating that the system has reached saturated boiling. The heating medium temperature can then be reduced to maintain a gentle, simmering extraction state. Compared to traditional temperature detection standards, this method identifies saturated boiling 8.5 minutes earlier, shortening the heating phase of the extraction process by 30%, thus improving production efficiency. The reduced heating time also decreases heating during the extraction process, contributing to energy savings.

[0062] The embodiments described above provide a detailed explanation of the technical solutions of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the present invention. Any modifications, additions, or similar substitutions made within the scope of the principles of the present invention should be included within the protection scope of the present invention.

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: Principal component analysis is performed on the first to the N-1th acoustic emission spectrum obtained in step 1 to establish a first PCA model and determine a 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 At this point, it is judged to be in a boiling state; In steps 2 and 4, the power spectral density within the 75-100 kHz 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; 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 observed value deviates significantly from the central location of other observed values. A Indicates the number of principal components retained; Official (1); In formula (1), For the nth sampling point A A vector composed of the principal component scores. It is a vector composed of the average scores of each principal component across all sampling points. yes A The diagonal matrix formed by the eigenvalues ​​corresponding to the principal components; Hotelling's T 2 Batch control limits for statistics The F-distribution obtained using formula (2) is as follows: Official (2); A This indicates the number of principal components retained. N Indicates the total number of sampling points. Is it subject to the degree of freedom? The F-distribution is at the upper critical value of the significance level α.

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, 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.

5. 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 1-4.

6. 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 1-4 through the computer program.

Images