Chaotic optimization model-based traditional Chinese medicine substance extraction process enhancement method and system

Abstract

The invention discloses a chaos optimization model-based traditional Chinese medicine substance extraction process enhancement method and system, and the method comprises the steps: selecting a substance extraction medicine or medicinal material, designing an experiment, carrying out the extraction of traditional Chinese medicine substances according to the extraction conditions given by the experiment to obtain an extraction solution, and measuring the content of an extracted substance in the extraction solution; inputting the extraction condition into a chaos optimization model, optimizing the extraction condition through an initial chaos model in the chaos optimization model, and inputting the optimized extraction condition into a neural network prediction model to obtain the content of an extracted substance; verifying the extraction condition, and if the verification is not true, carrying out the experiment again according to the optimized extraction condition given by the initial chaos model; and testing the stability of the traditional Chinese medicine substance extraction process scheme containing the optimized extraction conditions. According to the method, a chaos optimization model is established to quantitatively describe a complex nonlinear relationship between an extraction process and traditional Chinese medicine components, and optimal process conditions are optimized, so that a whole set of relatively complete optimal theoretical analysis system is established.

Description

Technical Field

[0001] The present invention belongs to the technical field of traditional Chinese medicine material extraction process optimization, and specifically relates to a traditional Chinese medicine material extraction process enhancement method and system based on a chaos optimization model. Background Art

[0002] The research on chemical substances and the basic research on pharmacological effects of traditional Chinese medicine is a complex and huge project, involving the intersection and integration of multiple disciplines such as traditional Chinese medicine, pharmacology, mathematics and molecular biology. Related laboratory research and clinical research have generated various types of data, which have two typical characteristics: one is that there are mutual influences between the data (synergistic / antagonistic effects), and the other is that the data types are complex. Therefore, the analysis of traditional Chinese medicine extraction data needs to comprehensively consider the integrity of the prescription and process, not only to find effective chemical components, but also to conduct in-depth research on the combination, quantification and interaction of different chemical components.

[0003] In the optimization of Chinese medicine extraction process, traditional methods such as orthogonal experimental design, central experimental design and uniform experimental design can infer the overall best experimental scheme through limited experimental data, but they have limitations. The optimal values ​​obtained by these methods are limited to a certain combination of the levels used in the experiment, cannot exceed the range of the levels taken, and lack clear directionality. Although response surface analysis can solve multivariate problems to a certain extent, its performance is not ideal when there is a highly nonlinear relationship between process conditions.

[0004] In recent years, the introduction of machine learning methods has provided new ideas for the optimization of Chinese medicine extraction technology. The combination of BP neural network and genetic algorithm has been widely used in the research of Chinese medicine extraction technology. However, BP neural network has problems such as overfitting, falling into local extreme values ​​and initial weight sensitivity, which leads to low accuracy of the model on the test set and different results may be obtained in each training. In order to overcome these shortcomings, global optimization algorithms such as taboo search, simulated annealing, genetic algorithm, ant colony algorithm and particle swarm algorithm have been applied to the optimization of Chinese medicine extraction technology. However, the determination of parameters in these algorithms often relies on practical experience, which brings great blindness and inconvenience to the teachers, students and scientific researchers of Chinese medicine. As a novel optimization technology, chaos optimization algorithm uses the characteristics of chaotic orbit traversal to effectively avoid falling into local minimum in the search process, so as to find the global optimal value. However, the research and application of chaos optimization neural network in the optimization of Chinese medicine extraction technology is relatively small and little is known. Although some scholars have conducted relevant research on the extraction process of some Chinese medicines and established BP neural network models and optimization algorithms, such methods have not yet gotten rid of the limitations of BP neural network.

[0005] In addition, in the multi-objective optimization of traditional Chinese medicine extraction technology, the assignment of weights becomes a key issue. Although subjective weighting methods such as the Delphi method and the analytic hierarchy process can give weights that meet the subject theory and practical needs based on the experience of the experimenter, there is a certain degree of subjective arbitrariness. The entropy weight method, as an objective weighting method, can be based on data-driven weighting and avoid the subjectivity of weight assignment, but its application in the optimization of traditional Chinese medicine extraction technology still needs further exploration.

[0006] In terms of software development, although there are professional software such as Orthogonal Analysis Assistant, Design-Expert, SAS, SPSS, and the neural network toolbox in Matlab that can be used for processing and analyzing TCM experimental data and clinical data, when the research problem requires the combination of machine learning models and chaos optimization algorithms, the use of these software still has certain difficulties. At the same time, there is currently no professional software based on machine learning models and chaos optimization algorithms for the optimization of TCM extraction processes.

[0007] In summary, it is necessary to further research and develop new optimization technologies to improve the optimization efficiency and accuracy of traditional Chinese medicine extraction process. Summary of the invention

[0008] In view of the above, the purpose of the present invention is to provide a method and system for enhancing the extraction process of traditional Chinese medicine substances based on a chaotic optimization model. By establishing a chaotic optimization model to quantitatively characterize the complex nonlinear relationship between the extraction process and the traditional Chinese medicine components, and using a chaotic optimization algorithm to optimize the optimal process conditions, a relatively complete set of theoretical analysis systems for optimization is established, forming an innovative scientific research method, enriching research means, and serving the vast number of teachers, students and scientific researchers in the field of traditional Chinese medicine.

[0009] In order to achieve the above-mentioned invention object, the technical solution provided by the present invention is as follows: The embodiment of the present invention provides a method for enhancing the extraction process of traditional Chinese medicine substances based on a chaotic optimization model, comprising the following steps: Step 1: Select drugs or medicinal materials for substance extraction and design experiments, extract Chinese medicine substances according to the given extraction conditions of the experiment to obtain extracts, and measure the content of the extracted substances in the extracts; Step 2: Input the extraction conditions obtained in step 1 into the chaos optimization model, optimize the extraction conditions in the chaos optimization model through the initial chaos model, and then input them into the neural network prediction model to obtain the content of the extracted substance; Step 3: Verify the optimized extraction conditions obtained by the initial chaotic model in step 2. If the verification is not successful, conduct the experiment again according to the optimized extraction conditions given by the initial chaotic model, and repeat steps 1 and 2. Step 4: Test the stability of the extraction process of traditional Chinese medicine substances containing the optimized extraction conditions obtained in step 2 or step 3, and calculate the n The ratio of the average content of the extracted substance in each experiment to the standard deviation is used to judge the stability of the extraction process of traditional Chinese medicine substances.

[0010] Preferably, the specific steps of step 2 include: Step 2a: Determine the extraction conditions, optimize the extraction conditions through the initial chaotic model and input them into the neural network prediction model as independent variables, and obtain the content of the extracted substances under different extraction conditions as the dependent variable output; Step 2b: Use the chaos optimization algorithm to optimize the neural network prediction model, use the material extraction content as the evaluation function, and continuously adjust the network weights to obtain the optimal neural network prediction model.

[0011] Preferably, the specific steps of step 2b include: Step 2b1: Set the number of iterations; Step 2b2: Obtain training set and validation set, and perform data normalization; Step 2b3: Determine the number of nodes in the input layer, hidden layer, and output layer of the neural network; Step 2b4: Initialize the weights of the input layer and the hidden layer using the chaos optimization algorithm, and initialize various operation parameters including the maximum number of iterations, the starting value and the end value of the algorithm search; Step 2b5: Use the substance extraction content as the evaluation function to determine whether the result converges. If it converges, output the optimal weight. If it does not converge, continue to search for the optimal solution through iteration.

[0012] Preferably, the convergence condition is that the number of iterations reaches the set maximum value, or the weight after multiple consecutive iterations of optimization does not differ from the current weight by more than a set weight threshold, or the extracted substance content meets a given accuracy requirement.

[0013] Preferably, in step 3, the condition for the verification to fail is that after multiple experiments, the ratio of the average content of the extracted substance to the standard deviation is less than a specified threshold.

[0014] Preferably, in step 4, by calculating n The ratio of the average content of the substance extracted in the experiment to the standard deviation, if the ratio is greater than the set threshold X , then the extraction process of traditional Chinese medicine substances is judged to be stable and the iteration is stopped; if the ratio is less than or equal to the set threshold X , it is judged as unstable and the iteration continues.

[0015] Preferably, the extraction conditions include extraction solvent, solvent concentration, extraction temperature, extraction time, solid-liquid ratio and extraction times.

[0016] Preferably, the initial chaos model adopts a cellular automaton chaos model, a Logistic chaos model or a Henon chaos model.

[0017] Preferably, the cellular automaton chaos model is selected from Moore-type, Elban-type or wandering-type cellular automaton models.

[0018] To achieve the same purpose of the invention, the embodiment of the present invention also provides a Chinese medicine material extraction process enhancement system based on a chaotic optimization model, which is implemented using the above-mentioned Chinese medicine material extraction process enhancement method based on a chaotic optimization model, and includes: an experimental design module, a chaotic optimization module, a condition verification module and a process evaluation module; The experimental design module is used to select drugs or medicinal materials for substance extraction and design experiments, extract Chinese medicine substances according to the extraction conditions given in the experiment to obtain extracts, and measure the content of the extracted substances in the extracts; The chaos optimization module is used to input the extraction conditions obtained in step 1 into the chaos optimization model, optimize the extraction conditions in the chaos optimization model through the initial chaos model, and then input them into the neural network prediction model to obtain the content of the extracted substance; The condition verification module is used to verify the optimized extraction conditions obtained by the initial chaotic model in step 2. If the verification is not established, the experiment is conducted again according to the optimized extraction conditions given by the initial chaotic model, and steps 1 and 2 are repeated; The process evaluation module is used to test the stability of the traditional Chinese medicine material extraction process scheme including the optimized extraction conditions obtained in step 2 or step 3, and calculate the n The ratio of the average content of the extracted substance in each experiment to the standard deviation is used to judge the stability of the extraction process of traditional Chinese medicine substances.

[0019] Compared with the prior art, the present invention has the following beneficial effects: The present invention applies a chaos optimization model to the extraction process of traditional Chinese medicine substances. By constructing a chaos optimization model, the extraction conditions are optimized and analyzed, and the optimal solution is obtained through an iterative algorithm. Finally, a verification experiment is performed to confirm whether the result is accurate, and the optimized extraction process is obtained. The content and accuracy of the extracted substances can be effectively improved, and the stability of the extracted substances is improved, thereby greatly improving the extraction efficiency and extraction rate, and supporting the extraction of multiple components. It has a wide range of uses and a flexible method, which helps to further reduce production costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0021] Figure 1 It is a schematic flow chart of a method for enhancing the extraction process of traditional Chinese medicine substances based on a chaotic optimization model provided by an embodiment of the present invention; Figure 2 It is a schematic diagram of the training process of the BP neural network model provided by an embodiment of the present invention; Figure 3 It is a schematic diagram of the overall framework of the main page of the optimization of the extraction process of traditional Chinese medicine designed in the software provided in the embodiment of the present invention; Figure 4 It is a structural schematic diagram of a Chinese medicine substance extraction process enhancement system based on a chaotic optimization model provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0022] To make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific implementation methods described herein are only used to explain the present invention and do not limit the scope of protection of the present invention.

[0023] The inventive concept of the present invention is: in view of the problems in the prior art that the machine learning methods such as the methods based on BP neural network and genetic algorithm have overfitting leading to insufficient accuracy, strong subjectivity in weight assignment, and lack of professional software based on machine learning models and chaos optimization algorithms, the embodiment of the present invention provides a method and system for enhancing the extraction process of Chinese medicine substances based on a chaos optimization model, combining BP neural network and chaos optimization algorithm, based on neural network theory, using chaos optimization algorithm to optimize network weights, establishing a chaos optimization model of Chinese medicine extraction process data, using the global optimization function of chaos optimization algorithm to realize autonomous search for the optimal extraction process combination, and experimentally verifying the predicted optimal process in the laboratory, analyzing the applicability of the model and the evaluation criteria. In addition, the graphical user interface development environment (GUIDE) in Matlab software, as well as related toolboxes and software packages are selected to realize the design of the extraction process parameter optimization interface, meet the actual needs of relevant professionals, and lay a foundation for promoting the combination of computer technology and Chinese medicine extraction technology, thereby promoting the sustainable development and utilization of Chinese medicine resources.

[0024] Figure 1 FIG. 1 is a flow chart of a method for enhancing the extraction process of traditional Chinese medicine substances based on a chaotic optimization model provided by an embodiment of the present invention. Figure 1 As shown, the embodiment provides a method for enhancing the extraction process of traditional Chinese medicine substances based on a chaotic optimization model, comprising the following steps: Step 1: Select drugs or medicinal materials for substance extraction and design experiments, extract Chinese medicine substances according to the given extraction conditions of the experiment to obtain extracts, and measure the content of extracted substances in the extracts.

[0025] In the embodiment, the selected medicinal material for substance extraction is Ziziphus jujuba seed, the content of the extracted substance is the total flavonoids, total saponins or total alkaloids, and the designed experimental content mainly includes the following aspects.

[0026] 1. Pretreatment of sour jujube kernels.

[0027] The preparation method of the fried jujube kernel defatted powder is as follows: the fried jujube kernel is crushed into fine powder, and the powder is screened with a 40-mesh sieve to make the particle size uniform. The powder is defatted with petroleum ether for 6 hours to remove the oil component in the powder. The defatted powder is placed in an oven and dried at a constant temperature of 50°C to obtain the fried jujube kernel defatted powder.

[0028] 2. Preparation of test solution.

[0029] Weigh 5.0g of fried jujube kernel defatted powder, extract with 60% ethanol for 30min at 70℃ and a solid-liquid ratio of 1:20 according to the reflux extraction method, recover the solvent to obtain the residue, dissolve it with 50mL of water, and then add water-saturated n-butanol to extract 3 times, adding 40mL of extractant each time. The n-butanol layers are combined, and the solvent is recovered to obtain the residue. The residue is dissolved with methanol and transferred to a 25mL brown volumetric flask and fixed to volume. After fixing to volume, draw 1mL of filtrate into a 10mL brown volumetric flask and fix to volume with CH3OH, mix well.

[0030] 3. Selection of measurement wavelength.

[0031] Take 200μL of the prepared test solution and transfer it to a 10mL stoppered test tube. After the solvent evaporates, take 200μL of 5% vanillin-glacial acetic acid solution and 800μL of HCIO4 and add them. After the mixture is sealed and mixed, place it in a water bath at 60℃ for 15min. After the temperature of the sample drops to room temperature, add 5mL of ice CH3COOH and mix. Use 200μL of methanol as a blank reagent in a 10mL stoppered test tube and perform the same sample preparation operation. The scanning range is 200~800nm, and its maximum absorption wavelength is determined as the measurement wavelength.

[0032] 4. Linearity inspection of standard curve and determination of total saponins and total flavonoids.

[0033] Accurately pipette 0 mL, 0.4 mL, 0.6 mL, 0.8 mL, 1.0 mL, and 1.2 mL of ginsenoside Re standard solution into a 10 mL stoppered colorimetric tube, place in a water bath to evaporate the solvent, add 0.2 mL of vanillin glacial acetic acid solution, then add 0.8 mL of perchloric acid, mix well, dissolve all the residue, place in a 60°C water bath and heat for 10 min, take out, cool in an ice bath, add 5.0 mL of glacial acetic acid, shake well, and immediately measure the absorbance at a wavelength of 560 nm. With absorbance A as the ordinate and saponin concentration C1 (mg / mL) as the abscissa, perform linear regression fitting calculation to obtain the standard curve of saponins. If If it is close to 1, it means that the data fit is high, and the data in the time period with a good linear relationship between the reference sample addition amount and the absorbance value are highly reliable and can be used for subsequent data calculations.

[0034] Take 10.0 mL of the total saponin extract of Ziziphus jujuba seed obtained by the experimental extraction, and determine the absorbance in the same way as the standard solution. Use a UV spectrophotometer to measure 5 times in parallel, and the calculation formula for the total saponin extraction amount X1 (mg / g) is: , in, is the content of total saponins in the sample (calculated as ginsenoside Re), in milligrams per gram (mg / g), The mass of ginsenoside Re in the test solution is calculated from the standard curve after sample background correction, in milligrams per milliliter (mg / mL); is the fixed volume of the sample solution to be measured, in milliliters (mL); is the color development volume of the sample solution to be measured, in milliliters (mL); is the weight of the sample, in grams (g).

[0035] Accurately weigh 10 mg of flavonoid standard in a 100 ml volumetric flask and dilute with 60% ethanol solution. Pipette 0, 2, 4, 6, 8, and 10 ml in a 50 ml volumetric flask, add ethanol to 10 ml, dilute with 0.2% sodium nitrite solution, place in a 50°C water bath for 30 minutes, cool to room temperature, and measure the absorbance at a wavelength of 292 nm with an ultraviolet spectrophotometer. With absorbance A as the ordinate and flavonoid concentration C2 (mg / 50ml) as the abscissa, perform linear regression fitting calculation to obtain the standard curve of flavonoids.

[0036] Take 10 ml of ethanol in a 50 ml volumetric flask, add 0.2% sodium nitrite solution to the mark to prepare a blank solution. Take the total flavonoids extract of Ziziphus jujuba seeds obtained by experimental extraction, and measure it 5 times in parallel with an ultraviolet spectrophotometer. The calculation formula for the total flavonoids extraction amount M2 (mg / g) is: , Where C2 is the total flavonoids concentration (mg / 50ml), V is the volume of the sample solution to be measured (ml), N is the dilution factor, W The dosage of raw material Chinese jujube kernel (g).

[0037] 5. Methodological investigation.

[0038] (1) Precision investigation: Accurately measure 0.2 mL of the reference solution into a 10 mL stoppered test tube, measure the absorbance after color development, measure the absorbance six times continuously, and calculate the RSD from the results to verify that the instrument has good precision.

[0039] (2) Stability test: Take the same test solution, measure the absorbance every 30 minutes from 0h to 3h after color development, and calculate the RSD from the results to verify that the test sample solution is stable within 3h.

[0040] (3) Repeatability study: Weigh 6 portions of medicinal materials from the same batch and prepare them in the same way as above. After color development, the absorbance is measured and the RSD is calculated from the results to verify that this method has good repeatability.

[0041] (4) Sample recovery investigation: Take 6 portions of crude powder of medicinal materials with known content, prepare them as above, accurately add appropriate amount of reference substance, use the same sample preparation and color development methods as above, measure the absorbance, and calculate the average recovery rate and RSD value.

[0042] 6. Single factor investigation.

[0043] (1) Investigation of solvent concentration: Take about 5.0g of dried powder of Semen Ziziphi Spinosae, weigh it accurately, and reflux extract it. The solvent concentration is set in a gradient of 10% every time, from 60% to 100%, with a total of 5 parallel groups. The other conditions are prepared according to the "test solution preparation" method. According to the results, a graph of the change of total saponin and total flavonoid content with the influence of solvent concentration is drawn. According to the solvent concentration range with the largest change slope shown in the graph, the subsequent response surface experiment extraction solvent concentration level is selected.

[0044] (2) Investigation of extraction temperature: Take about 5.0g of dried powder of Semen Ziziphi Spinosae, weigh it accurately, and reflux extract it. The temperature is set in a gradient of 10℃, from 50℃ to 90℃, with a total of 5 parallel groups. The other conditions are prepared according to the "Preparation of Test Solution" method. According to the results, a graph of the change of total saponin and total flavonoid content with the influence of extraction temperature is drawn. According to the temperature range with the largest change slope shown in the graph, the extraction temperature level of the subsequent response surface experiment is selected.

[0045] (3) Investigation of extraction time: Take about 5.0g of dried powder of Semen Ziziphi Spinosae, weigh it accurately, and reflux extract it with 60% ethanol as solvent. The time is set at a gradient interval of 30 min, with 4 parallel groups from 30min to 120min. The other conditions are prepared according to the "test solution preparation" method. According to the results, a graph of the change of total saponin and total flavonoid content with the effect of extraction time is drawn. According to the extraction time range with the largest slope shown in the graph, the extraction time level of the subsequent response surface experiment is selected.

[0046] (4) Investigation of material-liquid ratio: Take about 5.0g of dried powder of Chinese date seed, weigh it accurately, and reflux extract it. The solid-liquid ratio is set in a gradient of 10 times the amount of solvent, from 1:10 to 1:50, with a total of 5 parallel groups. The other conditions are prepared according to the "test solution preparation" method. According to the results, the total saponin and total flavonoid content changes with the solid-liquid ratio are plotted. According to the solid-liquid ratio range with the largest change slope shown in the figure, the subsequent response surface test solid-liquid ratio level is selected.

[0047] 7. Response surface experimental design.

[0048] According to the results of the single factor test, a total of four influencing factors were set, namely solvent concentration (A), extraction temperature (B), solid-liquid ratio (C) and extraction time (D). The content of total saponins was set as the evaluation index, and the response surface test was carried out using a four-factor three-level response surface design table. Finally, the total saponin extraction rate was calculated and the results were analyzed. According to the size of the range R, it can be concluded that the order of the influence of different factors on the saponin yield and the greatest influence of each factor at different levels can be obtained, so the optimal extraction process can be analyzed.

[0049] 8. Calculation of comprehensive evaluation value.

[0050] The entropy weight method was used to calculate the weight values, and the weight coefficients of total saponins and total flavonoids in Ziziphus jujuba seed were m and n . Comprehensive evaluation value = saponin content × m +Total flavonoids content of licorice× n .

[0051] 9. Verification test.

[0052] The variance and regression analysis of 30 groups of response surface experimental results were performed to obtain the extraction and purification process conditions. The optimal extraction process obtained by the response surface experiment was used for 3 extractions, and the test solution was prepared according to the above method and the color was developed using a color developer. After the absorbance was measured, the 3 extraction rates were calculated by the standard curve to verify whether the comprehensive evaluation value of the yield of the effective part of the Ziziphus jujuba seeds of the extraction process was relatively stable and whether it could be used as the optimal extraction process.

[0053] Step 2: Input the extraction conditions obtained in step 1 into the chaos optimization model. In the chaos optimization model, the extraction conditions are optimized by the initial chaos model and then input into the neural network prediction model to obtain the content of the extracted substance.

[0054] In the embodiment, the content of the extracted substance obtained in step 1 is used as the dependent variable, and the extraction conditions are used as the independent variable. The chaos optimization model is used to optimize the extraction conditions in step 1, and the content of the extracted substance under different extraction conditions is calculated. The specific steps are as follows.

[0055] Step 2a: Determine the extraction conditions (including that the extraction solvent is preferably selected from ethanol or other mixed solvents, the solvent concentration range is preferably from 40 to 95%, the extraction temperature range is preferably from 70 to 150°C, the extraction time range is preferably from 0.5 to 10h, the solid-liquid ratio range is preferably from 1 to 10, and the number of extractions range is preferably from 1 to 5), and optimize the extraction conditions through the initial chaos model and input them into the neural network prediction model as independent variables to obtain the content of the extracted substance under different extraction conditions as the dependent variable output. In the embodiment, the initial chaos model adopts a cellular automaton chaos model (Moore type, Elban type or wandering cellular automaton model), a Logistic chaos model or a Henon chaos model.

[0056] Step 2b: Use a chaos optimization algorithm to optimize the neural network prediction model, use the substance extraction content as an evaluation function, and continuously adjust the network weights to obtain the optimal neural network prediction model. In the embodiment, the neural network prediction model uses a BP neural network model.

[0057] like Figure 2 As shown, the training process of the BP neural network model is as follows: Step 2b1: Set the number of iterations.

[0058] Step 2b2: Get the training set and the validation set, and perform data normalization. The K-fold cross-validation method is used to randomly divide the data into a training set and a test set, and adjust the data order of the training set to a natural order.

[0059] Step 2b3: Determine the number of nodes in the input layer, hidden layer, and output layer of the neural network. The selection of the number of hidden layer nodes is a very complex issue, and there is currently no theoretical guidance. Many scholars have given various solutions to this problem, and have also proposed a number of empirical formulas. Previous studies have found that for different traditional Chinese medicines and extraction processes, the algorithms and empirical formulas for determining the number of hidden layer neurons are very different. By combining empirical formulas: , in, k is the sample size, n 1 is the number of neurons in the hidden layer, n is the number of nodes in the input layer, m is the number of nodes in the output layer. In the study of Chinese medicine extraction technology, a Represents an empirical adjustment parameter that affects the number of hidden layer neurons and is related to the complexity of the extraction process. Based on previous research and experimental experience, a The value range is from 0 to 5, and the specific value is determined by the experimental objectives and the properties of the extract. a The maximum value is 5, and the simultaneous formula is used to solve the problem, and an interval about the number of hidden layer nodes is obtained. This interval means an interval where the optimal number of hidden layer nodes is located. As long as you search in this interval, you can find it. Although this is a helpless method, it is effective and very suitable and practical for teachers, students and experimental personnel of traditional Chinese medicine.

[0060] Step 2b4: Use the chaos optimization algorithm to initialize the weights of the input layer and the hidden layer, and initialize various operation parameters including the maximum number of iterations, the starting value and the end value of the algorithm search.

[0061] Step 2b5: Taking the substance extraction content as the evaluation function, the evaluation function can be constructed based on factors such as the inverse number, fitting error and prediction error, and is used to optimize the weights between the layers of the BP neural network model, and determine whether the result converges. If it converges, the optimal weight is output, and if it does not converge, the optimal solution is continued to be found through iteration. Among them, the convergence condition is that the number of iterations reaches the set maximum value, or the weight after multiple consecutive iterations is optimized and the current weight does not differ by more than the set weight threshold, or the extracted substance content meets the given accuracy requirements.

[0062] In the embodiment, an evaluation function is constructed using the inverse number, the fitting error and the prediction error, and is used to optimize the weights between the layers of the BP neural network model.

[0063] First, define the inversion number: if the order of a pair of fitted data is opposite to that of the experimental data, they are called an inversion, and the sum of the inversions in the entire data is called the inversion number of the model, recorded as. For example, the order of the experimental data is 12345, and the order of the fitted data is 12354, then the inversion number of the model is ; The order of the fitted data is 54321, so the reverse order of the model is It can be seen that the smaller the inverse number of the model is, the more it can preserve the synergistic / antagonistic effects between the TCM extraction process conditions.

[0064] Next, define the normalized mean square error (NMSE) as: , in, The experimental data of the concentration of Chinese medicine extracts are is the average value of the experimental data, It is the predicted value of a data set (training set or test set) obtained from the model of the training set.

[0065] Finally, based on the theoretical knowledge of traditional Chinese medicine, the evaluation criteria for the mathematical model describing the optimal extraction process of traditional Chinese medicine are constructed while ensuring the stability of the model and not over-fitting: , in, After obtaining the content of the extracted substance through experiments, the error is calculated by fitting the experimental data using a mathematical model (such as linear regression or nonlinear fitting model). The calculation steps are as follows: 1. Collect experimental datasets ; 2. Use a fitting algorithm (such as the least squares method) to establish a fitting curve and obtain the fitting value ; 3. Calculate the error using the following formula: .

[0066] It is the error between the extracted substance content predicted by the BP neural network model and the actual experimental value. The calculation steps are as follows: 1. Train the BP neural network through experimental data and adjust the network weights and parameters; 2. Use an independent test dataset Input the model and get the predicted value ; 3. Calculate the error using the following formula: .

[0067] Weight coefficient , and It is based on an experimental data-driven method, obtained through multiple experimental optimizations, and aims to balance the model fitting error, prediction error, and inverse number. Impact on the extraction process of traditional Chinese medicine.

[0068] Step 3: Verify the optimized extraction conditions obtained by the initial chaotic model in step 2. If the verification is not successful, conduct the experiment again according to the optimized extraction conditions given by the initial chaotic model, and repeat steps 1 and 2.

[0069] In the embodiment, the optimized extraction conditions obtained by the initial chaotic model in step 2 are subjected to experimental extraction again to verify whether the optimal Semen Ziziphi Spinosae extraction process conditions given in the model are feasible. The experimental extraction rate after model calculation can be regarded as a further verification of the response surface test. According to the extraction rate of the extract after model calculation, it is judged whether the Semen Ziziphi Spinosae effective matter extraction rate is greater than the extraction rate of the response surface experimental design. The positive contribution of the mathematical algorithm to the improvement of the extraction rate can be further evaluated. The practicality, effectiveness and applicability of the model are verified and analyzed by experiment. If the verification is established, the preferred extraction conditions are retained. If the verification is not established, the extraction conditions are re-optimized. Wherein, the condition that the verification is not established is: after multiple experiments, the ratio of the average content of the extracted substance to the standard deviation obtained is less than the specified threshold value. In the embodiment, the threshold value is set to 2, which is an empirical value obtained based on the stability analysis of the content of the extracted substance in multiple experiments, to ensure the stability of the process, and can also be set according to the experiment or specific process requirements.

[0070] It is worth noting that for multi-objective optimization problems, we can combine data-driven methods and expert review methods to empower them and then transform them into single-objective optimization problems; we can also directly adopt multi-objective optimization methods, which only requires the comprehensive evaluation criteria of multiple objectives. Both of the above considerations can follow the technical route and research plan of the previous step to obtain a mathematical model that conforms to the theoretical knowledge of traditional Chinese medicine.

[0071] Step 4: Test the stability of the extraction process of traditional Chinese medicine substances containing the optimized extraction conditions obtained in step 2 or step 3, and calculate the n The ratio of the average content of the extracted substance in each experiment to the standard deviation is used to judge the stability of the extraction process of traditional Chinese medicine substances.

[0072] In the embodiment, by calculating n The ratio of the average content of the substance extracted in the experiment to the standard deviation, if the ratio is greater than the set threshold X , then the extraction process of traditional Chinese medicine substances is judged to be stable and the iteration is stopped; if the ratio is less than or equal to the set threshold X , it is judged as unstable and the iteration continues.

[0073] Based on the same inventive concept, Figure 4 As shown, an embodiment of the present invention further provides a TCM material extraction process enhancement system 400 based on a chaos optimization model, comprising: an experimental design module 410 , a chaos optimization module 420 , a condition verification module 430 and a process evaluation module 440 .

[0074] The experimental design module 410 is used to select drugs or medicinal materials for substance extraction and design experiments, extract traditional Chinese medicine substances according to the extraction conditions given in the experiment to obtain an extract, and measure the content of the extracted substances in the extract.

[0075] The chaos optimization module 420 is used to input the extraction conditions obtained in step 1 into the chaos optimization model, in which the extraction conditions are optimized by the initial chaos model and then input into the neural network prediction model to obtain the content of the extracted substance.

[0076] The condition verification module 430 is used to verify the optimized extraction conditions obtained by the initial chaotic model in step 2. If the verification is not successful, the experiment is conducted again according to the optimized extraction conditions given by the initial chaotic model, and steps 1 and 2 are repeated.

[0077] The process evaluation module 440 is used to test the stability of the traditional Chinese medicine material extraction process scheme including the optimized extraction conditions obtained in step 2 or step 3, and calculate the n The ratio of the average content of the extracted substance in each experiment to the standard deviation is used to judge the stability of the extraction process of traditional Chinese medicine substances.

[0078] It should be noted that the TCM material extraction process enhancement system based on the chaos optimization model provided in the above-mentioned embodiment belongs to the same inventive concept as the TCM material extraction process enhancement method based on the chaos optimization model. The specific implementation process is detailed in the embodiment of the TCM material extraction process enhancement method based on the chaos optimization model, which will not be repeated here.

[0079] In addition, the present invention is based on the above-mentioned embodiment to provide a method for enhancing the extraction process of Chinese medicine substances based on a chaotic optimization model. The graphical user interface development environment (GUIDE) in Matlab software, as well as related toolboxes and software packages are used to implement the design of the extraction process parameter optimization interface, and an intuitive and easy-to-operate visual interface is developed. Specifically, according to the process of optimizing the extraction process of Chinese medicine, the design interface includes a network training module, a simulation prediction module, and a parameter optimization module, and corresponding data input and output functions are implemented in each part. The overall framework of the main page of the optimization of the extraction process of Chinese medicine designed in the software is as follows: Figure 3As shown. In the network training module, the chaos optimization algorithm is used to realize the optimization process of the neural network, which mainly includes data preprocessing, network optimization and network maintenance functions. In the simulation prediction module, it mainly includes parameter input, network output and model evaluation functions. In the parameter optimization module, it mainly includes parameter range, process optimization and result output functions. On the operation interface, the user only needs to enter the corresponding parameters and click the button in the interface. Through the internal operation and calculation of the system, the corresponding calculation results can be obtained and the corresponding functions can be realized. At the same time, the user can explain and evaluate the optimization method based on the chaos optimization model based on the comparison with the existing models and algorithms, which effectively illustrates the effectiveness and applicability of the chaos optimization model.

[0080] The specific implementation methods described above provide a detailed description of the technical solutions and beneficial effects of the present invention. It should be understood that the above is only the most preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, supplements and equivalent substitutions made within the scope of the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for enhancing the extraction process of traditional Chinese medicine substances based on a chaotic optimization model, characterized in that: The following steps are involved: Step 1: Select drugs or medicinal materials for substance extraction and design experiments, extract Chinese medicine substances according to the given extraction conditions of the experiment to obtain extracts, and measure the content of the extracted substances in the extracts; Step 2: Input the extraction conditions obtained in step 1 into the chaos optimization model, optimize the extraction conditions in the chaos optimization model through the initial chaos model, and then input them into the neural network prediction model to obtain the content of the extracted substance; Step 3: Verify the optimized extraction conditions obtained by the initial chaotic model in step 2. If the verification is not successful, conduct the experiment again according to the optimized extraction conditions given by the initial chaotic model, and repeat steps 1 and 2. Step 4: Test the stability of the extraction process of traditional Chinese medicine substances containing the optimized extraction conditions obtained in step 2 or step 3, and calculate the n The ratio of the average content of the extracted substance in each experiment to the standard deviation is used to judge the stability of the extraction process of traditional Chinese medicine substances.

2. The method for enhancing the extraction process of traditional Chinese medicine substances based on a chaotic optimization model according to claim 1, characterized in that: The specific steps of step 2 include: Step 2a: Determine the extraction conditions, optimize the extraction conditions through the initial chaotic model and input them into the neural network prediction model as independent variables, and obtain the content of the extracted substances under different extraction conditions as the dependent variable output; Step 2b: Use the chaos optimization algorithm to optimize the neural network prediction model, use the material extraction content as the evaluation function, and continuously adjust the network weights to obtain the optimal neural network prediction model.

3. The method for enhancing the extraction process of traditional Chinese medicine substances based on a chaotic optimization model according to claim 2, characterized in that: The specific steps of step 2b include: Step 2b1: Set the number of iterations; Step 2b2: Obtain training set and validation set, and perform data normalization; Step 2b3: Determine the number of nodes in the input layer, hidden layer, and output layer of the neural network; Step 2b4: Initialize the weights of the input layer and the hidden layer using the chaos optimization algorithm, and initialize various operation parameters including the maximum number of iterations, the starting value and the end value of the algorithm search; Step 2b5: Use the substance extraction content as the evaluation function to determine whether the result converges. If it converges, output the optimal weight. If it does not converge, continue to search for the optimal solution through iteration.

4. The method for enhancing the extraction process of traditional Chinese medicine substances based on a chaotic optimization model according to claim 3, characterized in that: The convergence condition is that the number of iterations reaches the set maximum value, or the weight after multiple consecutive iterations of optimization does not differ from the current weight by more than the set weight threshold, or the extracted material content meets the given accuracy requirements.

5. The method for enhancing the extraction process of traditional Chinese medicine substances based on a chaotic optimization model according to claim 1, characterized in that: In step 3, the condition for the verification to fail is that after multiple experiments, the ratio of the average content of the extracted substance to the standard deviation is less than the specified threshold.

6. The method for enhancing the extraction process of traditional Chinese medicine substances based on a chaotic optimization model according to claim 2, characterized in that: In step 4, by calculating n The ratio of the average content of the substance extracted in the experiment to the standard deviation, if the ratio is greater than the set threshold X , then the extraction process of traditional Chinese medicine substances is judged to be stable and the iteration is stopped; if the ratio is less than or equal to the set threshold X , it is judged as unstable and the iteration continues.

7. The method for enhancing the extraction process of traditional Chinese medicine substances based on a chaotic optimization model according to claim 1, characterized in that: The extraction conditions include extraction solvent, solvent concentration, extraction temperature, extraction time, solid-liquid ratio and extraction times.

8. The method for enhancing the extraction process of traditional Chinese medicine substances based on a chaotic optimization model according to claim 1 or 2, characterized in that: The initial chaos model adopts the cellular automaton chaos model, the Logistic chaos model or the Henon chaos model.

9. The method for enhancing the extraction process of traditional Chinese medicine substances based on a chaotic optimization model according to claim 8, characterized in that: The cellular automaton chaos model is selected from Moore-type, Elban-type or wandering-type cellular automaton models.

10. A system for enhancing the extraction process of traditional Chinese medicine substances based on a chaotic optimization model, which is implemented by using the method for enhancing the extraction process of traditional Chinese medicine substances based on a chaotic optimization model according to any one of claims 1 to 9, characterized in that: include: Experimental design module, chaos optimization module, condition verification module and process evaluation module; The experimental design module is used to select drugs or medicinal materials for substance extraction and design experiments, extract Chinese medicine substances according to the extraction conditions given in the experiment to obtain extracts, and measure the content of the extracted substances in the extracts; The chaos optimization module is used to input the extraction conditions obtained in step 1 into the chaos optimization model, optimize the extraction conditions in the chaos optimization model through the initial chaos model, and then input them into the neural network prediction model to obtain the content of the extracted substance; The condition verification module is used to verify the optimized extraction conditions obtained by the initial chaotic model in step 2. If the verification is not established, the experiment is conducted again according to the optimized extraction conditions given by the initial chaotic model, and steps 1 and 2 are repeated; The process evaluation module is used to test the stability of the traditional Chinese medicine material extraction process scheme including the optimized extraction conditions obtained in step 2 or step 3, and calculate the n The ratio of the average content of the extracted substance in each experiment to the standard deviation is used to judge the stability of the extraction process of traditional Chinese medicine substances.

Images