A Traditional Chinese Medicine Extraction Control System and Method Based on AI Recognition

Abstract

This invention discloses an AI-based traditional Chinese medicine (TCM) extraction control system and method, belonging to the field of extraction control technology. The invention involves real-time collection of different batches of TCM samples placed within a TCM extraction area, and the installation of a spectral recognition device within this area. The device then collects spectral data from different batches of TCM samples in real time, and simultaneously aggregates the collected spectral data to construct a sample set. The spectral data in the constructed sample set is then processed using data processing methods. Based on the processed spectral data, spectral parameters during the TCM extraction process are determined through intelligent recognition. A TCM extraction model is then constructed based on these determined spectral parameters. Finally, machine learning is used to control the extraction process using the constructed TCM extraction model, thereby improving the accuracy of TCM extraction control.

Description

Technical Field

[0001] This invention relates to the field of extraction control technology, specifically to a traditional Chinese medicine extraction control system and method based on AI recognition. Background Technology

[0002] Traditional Chinese medicine (TCM), as an important component of TCM, is characterized by its synergistic multi-component, multi-target therapeutic mechanisms and precise formulation under TCM theory. However, the extraction of TCM components varies greatly depending on factors such as planting environment, cultivation techniques, harvesting methods, and subsequent processing techniques, posing a significant challenge to the control of TCM extraction.

[0003] Existing technology, such as the invention patent application CN119668094B, discloses a dynamic adjustment method and system for a traditional Chinese medicine extraction device. The method includes: acquiring parameters during the extraction process of the traditional Chinese medicine extraction device, dividing the parameters according to fixed time intervals, calculating the entropy value of the corresponding parameter at each time interval, evaluating the solvent flow rate adjustment space based on the entropy value change trend, and generating a dynamic adjustment evaluation result for the solvent flow rate. In this invention, by dynamically adjusting multiple key operational parameters in the traditional Chinese medicine extraction process, such as solvent flow rate, extraction time, and temperature, the accuracy and flexibility of the extraction process are significantly improved. In each step, the system acquires data through a real-time feedback mechanism and predicts future adjustment needs based on the entropy value change trend. This adjustment method based on real-time data avoids extraction under traditional fixed conditions. By real-time monitoring of the concentration of medicinal components, over-extraction or under-extraction is avoided, thereby maximizing the concentration of medicinal components.

[0004] As can be seen from the above schemes, current methods for controlling the extraction of medicinal materials often rely on calculating various parameters of the extraction solution to determine changes during the extraction process, neglecting the influence of the internal components of the medicinal materials on the extraction, which has certain limitations. Summary of the Invention

[0005] The purpose of this invention is to provide a control system and method for the extraction of traditional Chinese medicine based on AI recognition, which solves the problems existing in the background technology.

[0006] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: The present invention provides a method for controlling the extraction of traditional Chinese medicine based on AI recognition, specifically including the following steps:

[0007] S1. Collect different batches of Chinese medicine samples in real time and place them in the Chinese medicine extraction area, and install a spectral recognition device in the Chinese medicine extraction area;

[0008] S2. Based on the installed spectral recognition equipment, spectral data of different batches of Chinese medicine samples are collected in real time, and the collected spectral data are summarized to construct a sample set.

[0009] S3. Process the spectral data in the constructed sample set using data processing methods to obtain the processed spectral data;

[0010] S4. Based on the processed spectral data, the spectral parameters in the extraction process of traditional Chinese medicine are determined through intelligent recognition, and a traditional Chinese medicine extraction model is constructed based on the determined spectral parameters.

[0011] S5. Extraction control of the constructed Chinese herbal medicine extraction model is achieved through machine learning.

[0012] Preferably, the real-time collection of different batches of traditional Chinese medicine samples and their placement within the traditional Chinese medicine extraction area, along with the installation of a spectral recognition device within the extraction area, includes the following steps:

[0013] Collect Chinese medicine samples from different regions and batches, and randomly select 3 Chinese medicine samples from each region and batch. After drying each sample, grind it into powder, filter it through a sieve, and then compress it into tablets.

[0014] After being compressed into tablets, the tablets are placed in the Chinese medicine extraction area, and n detection points are evenly set in each tablet sample.

[0015] Install the corresponding components of the spectrometer based on the set detection points, with one detection point corresponding to one component;

[0016] The spectrometer's detection parameters are adjusted based on the set detection points, and each sample of the compressed sheet is measured based on the adjusted detection parameters.

[0017] Preferably, the method of using an installed spectral recognition device to collect spectral data from different batches of traditional Chinese medicine samples in real time, and simultaneously summarizing the collected spectral data to construct a sample set, includes the following steps:

[0018] Set the collection period set of the spectrometer ,in, Indicates the first Spectral data acquired within each acquisition cycle, Indicates the number of data collection cycles;

[0019] The collected spectral data are summarized to obtain a spectral data sample set. As shown below:

[0020] ;

[0021] in, Indicates the number of testing points. This indicates the time during the j-th acquisition period. Spectral data collected at each detection point.

[0022] Preferably, the step of processing the spectral data in the constructed sample set to obtain the processed spectral data includes the following steps:

[0023] The average value of all spectral data is calculated using the mean algorithm, and the calculated average value is used as the average spectrum.

[0024] ;

[0025] in, Indicates the first The average spectrum calculated from each detection point Indicates the kth period Spectral data collected at each detection point;

[0026] After obtaining the average spectrum, the spectral data of each detection point in the sample are linearly regressed with the average spectrum to obtain the baseline shift and offset of each sample.

[0027] Based on the baseline shift and offset of each sample, multi-source scattering correction is performed on the spectral data of the samples;

[0028] The formula for multi-source scattering correction is shown below:

[0029]

[0030] in, This represents the corrected spectral data. This represents the baseline shift of the spectral data for the nth sample. This represents the offset of the spectral data of the nth sample.

[0031] Preferably, the step of determining the spectral parameters in the traditional Chinese medicine extraction process based on the processed spectral data through intelligent recognition, and constructing a traditional Chinese medicine extraction model based on the determined spectral parameters, includes the following steps:

[0032] S41. Perform intelligent identification on the processed spectral data to determine the changing characteristics of spectral data during the extraction of traditional Chinese medicine.

[0033] S42. Construct a traditional Chinese medicine extraction model based on the changing characteristics of spectral data and the mass transfer phenomenon during the extraction process.

[0034] Preferably, the step of intelligently identifying the changes in spectral data during the extraction of traditional Chinese medicine includes the following steps:

[0035] The content of each component in the processed spectral data was determined based on the relationship between the components of traditional Chinese medicine and spectral data.

[0036] The formulas for calculating the content of each component in the spectral data are shown below:

[0037] ;

[0038] in, This indicates the content of each component in the spectral data. Indicates spectral intensity, These are constants set for the spectrometer;

[0039] The processed spectral data are classified based on the component content, and each category corresponds to a set of components.

[0040] The spectral data of each component of different samples in each period are summarized to obtain the summarized spectral data, and the summarized spectral data is then input into a convolutional neural network for feature extraction.

[0041] Based on the input aggregated spectral data, the variation features of the aggregated spectral data are extracted and saved through the convolution operation of a convolutional neural network.

[0042] The formula for calculating convolution is as follows:

[0043] ;

[0044] in, This represents the aggregated spectral data input. represents the weights of the corresponding convolution kernel, and b represents the bias value. This indicates the changes in the extracted and summarized spectral data.

[0045] Preferably, the construction of the traditional Chinese medicine extraction model based on the changing characteristics of spectral data and the mass transfer phenomenon during the extraction process includes the following steps:

[0046] Based on the mass transfer phenomenon, the extraction process of traditional Chinese medicine is divided into internal diffusion process and external diffusion process.

[0047] The mass transfer differential equation for the external diffusion process is shown below:

[0048] ;

[0049] in, This represents the diffusion flux of the index component. Indicates the diffusion area. This represents the diffusion coefficient of the corresponding component. This represents the mass concentration gradient of the index component in the z-direction. Indicates the first Group index components;

[0050] The mass transfer differential equation for the internal diffusion process is shown below:

[0051] ;

[0052] Where t represents time;

[0053] The change characteristics of spectral data during the extraction process of traditional Chinese medicine are defined as the extraction rate of the extraction process. Based on the extraction rate and mass transfer phenomenon, a traditional Chinese medicine extraction model is constructed.

[0054] The extraction model of traditional Chinese medicine is shown below:

[0055] ;

[0056] in, These represent the extraction rates corresponding to the external diffusion process and the internal diffusion process, respectively. Indicates the extraction rate. Indicates the diffusion index, This indicates the amount extracted.

[0057] Preferably, the extraction control of the constructed traditional Chinese medicine extraction model using machine learning includes the following steps:

[0058] S51. Construct a control decision system for traditional Chinese medicine extraction based on the extraction model, including:

[0059] The cost control decision for traditional Chinese medicine extraction is calculated using economic analysis, and is expressed as follows:

[0060] ;

[0061] in, This indicates the change in revenue from the extraction of traditional Chinese medicine, expressed in yuan. This indicates the profit from the extraction of traditional Chinese medicine, expressed in yuan / kg. This indicates the cost of raw materials and extraction, in yuan. This parameter represents the relationship between extraction volume and extraction cost, and is set to be directly proportional to extraction volume and extraction cost.

[0062] S52. Based on the construction of a decision-making mechanism for the extraction control of traditional Chinese medicine, the extraction control of traditional Chinese medicine is carried out through machine learning.

[0063] Preferably, the step of controlling the extraction of traditional Chinese medicine (TCM) through machine learning based on the construction of TCM extraction control decisions includes the following steps:

[0064] S521. Summarize the processed spectral data and initialize the parameters:

[0065] Each particle represents a cost control decision for traditional Chinese medicine extraction based on a set of processed spectral data. The population size and maximum number of iterations are also defined. ;

[0066] S522. Calculate the fitness of each particle:

[0067] The cost control decision for traditional Chinese medicine extraction is used as the fitness function, and the fitness of each particle is calculated. The fitness calculation formula is shown below:

[0068] ;

[0069] in, Indicates particle fitness;

[0070] S523. Set a fitness threshold, and based on the calculated particle fitness and the set fitness threshold, select particles with fitness values ​​greater than the fitness threshold to construct a set of choices for the cost control of traditional Chinese medicine extraction.

[0071] S524. Update the particle positions in the decision selection set for cost control of traditional Chinese medicine extraction using an optimization algorithm;

[0072] The particle position update formula is as follows:

[0073] ;

[0074] in, Represents particles In the Position in the next iteration;

[0075] For each particle being computed, its fitness at its current position is compared with the best position it has passed. The fitness of the current position is compared with that of the best position it has visited. Based on the fitness level, the current position is taken as the current optimal position. If the fitness of the current position is less than or equal to the best position it has visited, The fitness of the position does not change the current optimal position. ;

[0076] S525. Sort the populations in descending order based on their fitness, and consider the bottom 25% of the populations as invalid populations. Delete the invalid populations and update the populations again.

[0077] S526. Repeat steps S522-S525 until the maximum number of iterations is reached, and output the Chinese medicine extraction cost control decision corresponding to the processed spectral data of the updated particle positions.

[0078] S527. Based on the output spectral data, the cost control decision for traditional Chinese medicine extraction is carried out to control the extraction of traditional Chinese medicine.

[0079] The present invention also provides an AI-based traditional Chinese medicine extraction control system, which is used to implement an AI-based traditional Chinese medicine extraction control method. The system includes: a data acquisition module, a data processing module, an intelligent recognition module, a model building module, and an extraction control module.

[0080] The data acquisition module is used to collect spectral data of different batches of traditional Chinese medicine samples at different periods in real time.

[0081] The data processing module is used to process the real-time acquired spectral data to obtain processed spectral data;

[0082] The intelligent recognition module is used to determine the spectral parameters in the extraction process of traditional Chinese medicine based on the processed spectral data;

[0083] The model building module is used to build a traditional Chinese medicine extraction model based on the spectral parameters determined during the extraction process.

[0084] The extraction control module is used to control the extraction of traditional Chinese medicine based on the constructed extraction model and extraction cost decision.

[0085] The beneficial effects of this invention are as follows:

[0086] (1) This invention collects different batches of Chinese medicine samples in real time and places them in the Chinese medicine extraction area. A spectral recognition device is installed in the Chinese medicine extraction area. Then, the spectral data of different batches of Chinese medicine samples are collected in real time based on the installed spectral recognition device. At the same time, the collected spectral data are summarized to construct a sample set. Then, the spectral data in the constructed sample set is processed by data processing. Based on the processed spectral data, the spectral parameters in the Chinese medicine extraction process are determined by intelligent recognition. Based on the determined spectral parameters, a Chinese medicine extraction model is constructed. Finally, the Chinese medicine extraction model is extracted and controlled by machine learning, which improves the accuracy of Chinese medicine extraction control.

[0087] (2) The present invention ensures the accuracy and reliability of the acquisition of spectral data of Chinese medicine samples by setting the acquisition cycle of the spectrometer and setting the detection points in the Chinese medicine samples.

[0088] (3) The present invention processes the spectral data in the constructed sample set through data processing, and adjusts the collected spectral data through data mean calculation and data correction, thereby improving the rationality of spectral data processing.

[0089] (4) This invention intelligently identifies the changes in spectral data during the extraction of Chinese medicine by performing intelligent identification on the processed spectral data. At the same time, it constructs a Chinese medicine extraction model based on the changes in spectral data and the mass transfer phenomenon during the extraction of Chinese medicine, thereby quantitatively monitoring the extraction process of Chinese medicine and improving the accuracy of Chinese medicine extraction.

[0090] (5) This invention trains and optimizes the constructed Chinese medicine extraction model by setting the Chinese medicine extraction control decision, and improves the real-time performance of Chinese medicine extraction control by using machine learning. Attached Figure Description

[0091] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0092] Figure 1 This is a schematic diagram of the process for controlling the extraction of traditional Chinese medicine according to the present invention. Detailed Implementation

[0093] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0094] In a specific embodiment of the present invention,

[0095] Reference Figure 1 As shown, this invention provides a method for controlling the extraction of traditional Chinese medicine based on AI recognition, comprising the following steps:

[0096] S1. Collect different batches of Chinese medicine samples in real time and place them in the Chinese medicine extraction area, and install a spectral recognition device in the Chinese medicine extraction area;

[0097] S2. Based on the installed spectral recognition equipment, spectral data of different batches of Chinese medicine samples are collected in real time, and the collected spectral data are summarized to construct a sample set.

[0098] S3. Process the spectral data in the constructed sample set using data processing methods to obtain the processed spectral data;

[0099] S4. Based on the processed spectral data, the spectral parameters in the extraction process of traditional Chinese medicine are determined through intelligent recognition, and a traditional Chinese medicine extraction model is constructed based on the determined spectral parameters.

[0100] S5. Extraction control of the constructed traditional Chinese medicine extraction model is achieved through machine learning.

[0101] Furthermore, referring to Figure 1As shown, the process of collecting different batches of traditional Chinese medicine samples in real time and placing them in the traditional Chinese medicine extraction area, and installing spectral recognition equipment in the traditional Chinese medicine extraction area includes the following steps:

[0102] Collect Chinese medicine samples from different regions and batches, and randomly select 3 Chinese medicine samples from each region and batch. After drying each sample, grind it into powder, filter it through a sieve, and then compress it into tablets.

[0103] Furthermore, after being compressed into tablets, the tablets are placed in the Chinese medicine extraction area, and n detection points are evenly set in each tablet sample;

[0104] Furthermore, the corresponding components of the spectrometer are installed based on the set detection points, with one detection point corresponding to one component;

[0105] Furthermore, the detection parameters of the spectrometer are adjusted based on the set detection points, and each sample of the compressed sheet is measured based on the adjusted detection parameters;

[0106] Furthermore, referring to Figure 1 As shown, the process of collecting spectral data from different batches of traditional Chinese medicine samples in real time using installed spectral recognition equipment, and simultaneously summarizing the collected spectral data to construct a sample set includes the following steps:

[0107] Set the collection period set of the spectrometer ,in, Indicates the first Spectral data acquired within each acquisition cycle, Indicates the number of data collection cycles;

[0108] The collected spectral data are summarized to obtain a spectral data sample set. As shown below:

[0109] ;

[0110] in, Indicates the number of testing points. This indicates the time during the j-th acquisition period. Spectral data collected at each detection point;

[0111] Furthermore, referring to Figure 1 As shown, the spectral data in the constructed sample set is processed using data processing methods to obtain the processed spectral data, including the following steps:

[0112] The average value of all spectral data is calculated using the mean algorithm, and the calculated average value is used as the average spectrum.

[0113] ;

[0114] in, Indicates the first The average spectrum calculated from each detection point Indicates the kth period Spectral data collected at each detection point;

[0115] Furthermore, after obtaining the average spectrum, the spectral data of each detection point in the sample are linearly regressed with the average spectrum to obtain the baseline shift and offset of each sample;

[0116] Furthermore, based on the baseline shift and offset of each sample, multi-source scattering correction is performed on the spectral data of the samples;

[0117] The formula for multi-source scattering correction is shown below:

[0118]

[0119] in, This represents the corrected spectral data. This represents the baseline shift of the spectral data for the nth sample. This represents the offset of the spectral data for the nth sample;

[0120] Furthermore, referring to Figure 1 As shown, based on the processed spectral data, the spectral parameters in the extraction process of traditional Chinese medicine are determined through intelligent recognition. Simultaneously, based on the determined spectral parameters, a traditional Chinese medicine extraction model is constructed, including the following steps:

[0121] S41. Perform intelligent identification on the processed spectral data to determine the changing characteristics of spectral data during the extraction of traditional Chinese medicine.

[0122] The content of each component in the processed spectral data was determined based on the relationship between the components of traditional Chinese medicine and spectral data.

[0123] The formulas for calculating the content of each component in the spectral data are shown below:

[0124] ;

[0125] in, This indicates the content of each component in the spectral data. Indicates spectral intensity, These are constants set for the spectrometer;

[0126] Furthermore, the processed spectral data are classified based on the component content, and a set of components is set for each category;

[0127] Furthermore, the spectral data corresponding to each component of different samples in each period are summarized to obtain the summarized spectral data, and the summarized spectral data is input into a convolutional neural network for feature extraction;

[0128] Based on the input aggregated spectral data, the variation features of the aggregated spectral data are extracted and saved through the convolution operation of a convolutional neural network.

[0129] The formula for calculating convolution is as follows:

[0130] ;

[0131] in, This represents the aggregated spectral data input. represents the weights of the corresponding convolution kernel, and b represents the bias value. This indicates the changes in the extracted and summarized spectral data;

[0132] S42. Construct a traditional Chinese medicine extraction model based on the changing characteristics of spectral data and the mass transfer phenomenon during the extraction process.

[0133] Based on the mass transfer phenomenon, the extraction process of traditional Chinese medicine is divided into internal diffusion process and external diffusion process.

[0134] The mass transfer differential equation for the external diffusion process is shown below:

[0135] ;

[0136] in, This represents the diffusion flux of the index component. Indicates the diffusion area. This represents the diffusion coefficient of the corresponding component. This represents the mass concentration gradient of the index component in the z-direction. Indicates the first Group index components;

[0137] The mass transfer differential equation for the internal diffusion process is shown below:

[0138] ;

[0139] Where t represents time;

[0140] Furthermore, the change characteristics of spectral data during the extraction process of traditional Chinese medicine are set as the extraction rate of the extraction process, and a traditional Chinese medicine extraction model is constructed based on the extraction rate and mass transfer phenomenon.

[0141] The extraction model of traditional Chinese medicine is shown below:

[0142] ;

[0143] in, These represent the extraction rates corresponding to the external diffusion process and the internal diffusion process, respectively. Indicates the extraction rate. Indicates the diffusion index, Indicates the amount extracted;

[0144] Furthermore, referring to Figure 1 As shown, the extraction control of the constructed traditional Chinese medicine extraction model using machine learning includes the following steps:

[0145] S51. Construct a control decision system for traditional Chinese medicine extraction based on the extraction model, including:

[0146] The cost control decision for traditional Chinese medicine extraction is calculated using economic analysis, and is expressed as follows:

[0147] ;

[0148] in, This indicates the change in revenue from the extraction of traditional Chinese medicine, expressed in yuan. This indicates the profit from the extraction of traditional Chinese medicine, expressed in yuan / kg. This indicates the cost of raw materials and extraction, in yuan. This parameter represents the relationship between extraction volume and extraction cost, and is set to be directly proportional to extraction volume and extraction cost.

[0149] S52. Based on the construction of a decision-making mechanism for the extraction control of traditional Chinese medicine, machine learning is used to control the extraction of traditional Chinese medicine.

[0150] S521. Summarize the processed spectral data and initialize the parameters:

[0151] Each particle represents a cost control decision for traditional Chinese medicine extraction based on a set of processed spectral data. The population size and maximum number of iterations are also defined. ;

[0152] S522. Calculate the fitness of each particle:

[0153] The cost control decision for traditional Chinese medicine extraction is used as the fitness function, and the fitness of each particle is calculated. The fitness calculation formula is shown below:

[0154] ;

[0155] in, Indicates particle fitness;

[0156] S523. Set a fitness threshold, and based on the calculated particle fitness and the set fitness threshold, select particles with fitness values ​​greater than the fitness threshold to construct a set of choices for the cost control of traditional Chinese medicine extraction.

[0157] S524. Update the particle positions in the decision selection set for cost control of traditional Chinese medicine extraction using an optimization algorithm;

[0158] The particle position update formula is as follows:

[0159] ;

[0160] in, Represents particles In the Position in the next iteration;

[0161] For each particle being computed, its fitness at its current position is compared with the best position it has passed. The fitness of the current position is compared with that of the best position it has visited. Based on the fitness level, the current position is taken as the current optimal position. If the fitness of the current position is less than or equal to the best position it has visited, The fitness of the position does not change the current optimal position. ;

[0162] S525. Sort the populations in descending order based on their fitness, and consider the bottom 25% of the populations as invalid populations. Delete the invalid populations and update the populations again.

[0163] S526. Repeat steps S522-S525 until the maximum number of iterations is reached, and output the Chinese medicine extraction cost control decision corresponding to the processed spectral data of the updated particle positions.

[0164] S527. Decision-making for cost control of traditional Chinese medicine extraction based on the output processed spectral data;

[0165] In one specific embodiment, the AI-based traditional Chinese medicine extraction control system is used to implement an AI-based traditional Chinese medicine extraction control method. The system includes: a data acquisition module, a data processing module, an intelligent recognition module, a model building module, and an extraction control module.

[0166] The data acquisition module is used to collect spectral data of different batches of traditional Chinese medicine samples at different periods in real time.

[0167] The data processing module is used to process the real-time acquired spectral data to obtain processed spectral data;

[0168] The intelligent recognition module is used to determine the spectral parameters in the extraction process of traditional Chinese medicine based on the processed spectral data;

[0169] The model building module is used to build a traditional Chinese medicine extraction model based on the spectral parameters determined during the extraction process.

[0170] The extraction control module is used to control the extraction of traditional Chinese medicine based on the constructed extraction model and extraction cost decision.

[0171] It should be noted that,

[0172] The above content is merely an example and illustration of the concept of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described, or use similar methods to replace them, as long as they do not deviate from the concept of the invention or exceed the scope defined by the present invention, and all such modifications and additions should fall within the protection scope of the present invention.

Claims

1. A method for controlling the extraction of traditional Chinese medicine based on AI recognition, characterized in that, Includes the following steps: S1. Collect different batches of Chinese medicine samples in real time and place them in the Chinese medicine extraction area, and install a spectral recognition device in the Chinese medicine extraction area; S2. Based on the installed spectral recognition equipment, spectral data of different batches of Chinese medicine samples are collected in real time, and the collected spectral data are summarized to construct a sample set. S3. Process the spectral data in the constructed sample set using data processing methods to obtain the processed spectral data; S4. Based on the processed spectral data, the spectral parameters in the extraction process of traditional Chinese medicine are determined through intelligent recognition, and a traditional Chinese medicine extraction model is constructed based on the determined spectral parameters. S41. Perform intelligent identification on the processed spectral data to determine the changing characteristics of spectral data during the extraction of traditional Chinese medicine. S42. Construct a traditional Chinese medicine extraction model based on the changing characteristics of spectral data and the mass transfer phenomenon during the extraction process. Based on the mass transfer phenomenon, the extraction process of traditional Chinese medicine is divided into internal diffusion process and external diffusion process. The mass transfer differential equation for the external diffusion process is shown below: ； in, This represents the diffusion flux of the index component. Indicates the diffusion area. This represents the diffusion coefficient of the corresponding component. This represents the mass concentration gradient of the index component in the z-direction. Indicates the first Group index components; The mass transfer differential equation for the internal diffusion process is shown below: ； Where t represents time; The change characteristics of spectral data during the extraction process of traditional Chinese medicine are defined as the extraction rate of the extraction process. Based on the extraction rate and mass transfer phenomenon, a traditional Chinese medicine extraction model is constructed. The extraction model of traditional Chinese medicine is shown below: ； in, These represent the extraction rates corresponding to the external diffusion process and the internal diffusion process, respectively. Indicates the extraction rate. Indicates the diffusion index. Indicates the amount extracted; S5. Extraction control of the constructed Chinese herbal medicine extraction model is achieved through machine learning.

2. The method for controlling the extraction of traditional Chinese medicine based on AI recognition according to claim 1, characterized in that, The real-time collection of different batches of traditional Chinese medicine samples and their placement within the traditional Chinese medicine extraction area, along with the installation of a spectral recognition device within the extraction area, includes the following steps: Collect Chinese medicine samples from different regions and batches, and randomly select 3 Chinese medicine samples from each region and batch. After drying each sample, grind it into powder, filter it through a sieve, and then compress it into tablets. After being compressed into tablets, the tablets are placed in the Chinese medicine extraction area, and n detection points are evenly set in each tablet sample. Install the corresponding components of the spectrometer based on the set detection points, with one detection point corresponding to one component; The spectrometer's detection parameters are adjusted based on the set detection points, and each sample of the compressed sheet is measured based on the adjusted detection parameters.

3. The method for controlling the extraction of traditional Chinese medicine based on AI recognition according to claim 1, characterized in that, The method of using installed spectral recognition equipment to collect spectral data from different batches of traditional Chinese medicine samples in real time, and then summarizing the collected spectral data to construct a sample set, includes the following steps: Set the collection period set of the spectrometer ,in, Indicates the first Spectral data acquired within each acquisition cycle, Indicates the number of data collection cycles; The collected spectral data are summarized to obtain a spectral data sample set. As shown below: ； in, Indicates the number of testing points. This indicates the time during the j-th acquisition period. Spectral data collected at each detection point.

4. The method for controlling the extraction of traditional Chinese medicine based on AI recognition according to claim 1, characterized in that, The process of processing the spectral data in the constructed sample set to obtain the processed spectral data includes the following steps: The average value of all spectral data is calculated using the mean algorithm, and the calculated average value is used as the average spectrum. ； in, Indicates the first The average spectrum calculated from each detection point Indicates the kth period Spectral data collected at each detection point; After obtaining the average spectrum, the spectral data of each detection point in the sample are linearly regressed with the average spectrum to obtain the baseline shift and offset of each sample. Based on the baseline shift and offset of each sample, multi-source scattering correction is performed on the spectral data of the samples; The formula for multi-source scattering correction is shown below: ； in, This represents the corrected spectral data. This represents the baseline shift of the spectral data for the nth sample. This represents the offset of the spectral data of the nth sample.

5. The method for controlling the extraction of traditional Chinese medicine based on AI recognition according to claim 1, characterized in that, The intelligent identification of the processed spectral data to determine the changing characteristics of the spectral data during the extraction of traditional Chinese medicine includes the following steps: The content of each component in the processed spectral data was determined based on the relationship between the components of traditional Chinese medicine and spectral data. The formulas for calculating the content of each component in the spectral data are shown below: ； in, This indicates the content of each component in the spectral data. Indicates spectral intensity, These are constants set for the spectrometer; The processed spectral data are classified based on the component content, and each category corresponds to a set of components. The spectral data of each component of different samples in each period are summarized to obtain the summarized spectral data, and the summarized spectral data is then input into a convolutional neural network for feature extraction. Based on the input aggregated spectral data, the variation features of the aggregated spectral data are extracted and saved through the convolution operation of a convolutional neural network. The formula for calculating convolution is as follows: ； in, This represents the aggregated spectral data input. represents the weights of the corresponding convolution kernel, and b represents the bias value. This indicates the changes in the extracted and summarized spectral data.

6. The method for controlling the extraction of traditional Chinese medicine based on AI recognition according to claim 1, characterized in that, The extraction control of the constructed traditional Chinese medicine extraction model using machine learning includes the following steps: S51. Construct a control decision system for traditional Chinese medicine extraction based on the extraction model, including: The cost control decision for traditional Chinese medicine extraction is calculated using economic analysis, and is expressed as follows: ； in, This indicates the change in revenue from the extraction of traditional Chinese medicine, expressed in yuan. This indicates the profit from the extraction of traditional Chinese medicine, expressed in yuan / kg. This indicates the cost of raw materials and extraction, in yuan. This parameter represents the relationship between extraction volume and extraction cost, and is set to be directly proportional to extraction volume and extraction cost. S52. Based on the construction of a decision-making mechanism for the extraction control of traditional Chinese medicine, the extraction control of traditional Chinese medicine is carried out through machine learning.

7. The method for controlling the extraction of traditional Chinese medicine based on AI recognition according to claim 6, characterized in that, The method of controlling the extraction of traditional Chinese medicine based on machine learning includes the following steps: S521. Summarize the processed spectral data and initialize the parameters: Each particle represents a cost control decision for traditional Chinese medicine extraction based on a set of processed spectral data. The population size and maximum number of iterations are also defined. ; S522. Calculate the fitness of each particle: The cost control decision for traditional Chinese medicine extraction is used as the fitness function, and the fitness of each particle is calculated. The fitness calculation formula is shown below: ； in, Indicates particle fitness; S523. Set a fitness threshold, and based on the calculated particle fitness and the set fitness threshold, select particles with fitness values ​​greater than the fitness threshold to construct a set of choices for the cost control of traditional Chinese medicine extraction. S524. Update the particle positions in the decision selection set for cost control of traditional Chinese medicine extraction using an optimization algorithm; The particle position update formula is as follows: ； in, Represents particles In the Position in the next iteration; For each particle being computed, its fitness at its current position is compared with the best position it has passed. The fitness of the current position is compared with that of the best position it has visited. Based on the fitness level, the current position is taken as the current optimal position. If the fitness of the current position is less than or equal to the best position it has visited, The fitness of the position does not change the current optimal position. ; S525. Sort the populations in descending order based on their fitness, and consider the bottom 25% of the populations as invalid populations. Delete the invalid populations and update the populations again. S526. Repeat steps S522-S525 until the maximum number of iterations is reached, and output the Chinese medicine extraction cost control decision corresponding to the processed spectral data of the updated particle positions. S527. Based on the output spectral data, the cost control decision for traditional Chinese medicine extraction is carried out to control the extraction of traditional Chinese medicine.

8. A system for implementing the AI-based identification-based traditional Chinese medicine extraction control method as described in claim 1, characterized in that, include: The system includes a data acquisition module, a data processing module, an intelligent recognition module, a model building module, and an extraction and control module. The data acquisition module is used to collect spectral data of different batches of traditional Chinese medicine samples at different periods in real time. The data processing module is used to process the real-time acquired spectral data to obtain processed spectral data; The intelligent recognition module is used to determine the spectral parameters in the extraction process of traditional Chinese medicine based on the processed spectral data; The model building module is used to build a traditional Chinese medicine extraction model based on the spectral parameters determined during the extraction process. The extraction control module is used to control the extraction of traditional Chinese medicine based on the constructed extraction model and extraction cost decision.

Images