An intelligent detection method for active ingredients of sleeping patch

Abstract

The present invention relates to the technical field of intelligent detection of chemical components, and in particular to an intelligent detection method for active components of a sleeping patch, comprising the following steps: step S1, establishing a feature extraction unit, for performing feature extraction on active components in the sleeping patch, and extracting feature information thereof; step S2, establishing a feature value comparison unit, comparing the feature information extracted by the feature extraction unit with the feature information of each active component, and obtaining the corresponding active component; step S3, establishing a feature value statistical unit, for statistically analyzing the proportions corresponding to the feature information of various active components in the same sample, and storing the statistical results in a database; step S4, establishing an active component analysis unit, for performing safety analysis on the active components in the sample, and at the same time reasonably allocating active components of different proportions. By establishing a multi-dimensional feature extraction and comparison mechanism, the method can accurately distinguish compounds with similar structures, and greatly reduce the situations of misjudgment and omission.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent detection of chemical components, and in particular to an intelligent detection method for effective components of a sleeping patch. Background Art

[0002] As an emerging non-drug sleep aid product, sleep patches have been widely popular in the market in recent years. Such products usually contain a variety of natural plant extracts, which play a role in helping sleep and improving sleep quality through transdermal absorption. With the rapid development of the sleep patch market, how to accurately and efficiently detect its active ingredients and ensure product quality and safety has become a major challenge facing the industry.

[0003] Traditional methods for detecting active ingredients in sleeping patches mainly rely on gas chromatography-mass spectrometry (GC-MS). Although this method can perform qualitative and quantitative analysis of the chemical components in the sample, it still has many limitations in practical applications. First, since sleeping patches often contain complex mixtures of plant extracts, traditional methods often misjudge or miss ingredients, especially for compounds with similar structures. Secondly, in terms of quantitative analysis, traditional methods mainly rely on peak area normalization, which is easily affected by factors such as baseline drift and peak overlap, resulting in inaccurate quantitative results.

[0004] In addition, existing detection methods are inefficient when processing large quantities of samples, and often require experienced operators to perform tedious manual data processing, which is not only time-consuming and labor-intensive, but also prone to subjective errors. More importantly, traditional safety assessment methods often only focus on whether the content of a single ingredient exceeds the standard, ignoring the possible synergistic or antagonistic effects between different ingredients, and are unable to fully evaluate the safety of the product.

[0005] In view of the above problems, it is urgent to develop an intelligent method that can quickly, accurately and comprehensively detect the active ingredients of sleep patches to meet the growing market demand and regulatory requirements. Summary of the invention

[0006] The present invention provides an intelligent detection method for the effective ingredients of a sleeping patch, aiming to solve the technical problems existing in the existing detection methods for the effective ingredients of sleeping patches, such as low recognition rate, insufficient quantitative accuracy, low detection efficiency and one-sided safety assessment.

[0007] The present invention provides a method for intelligently detecting active ingredients of a sleeping patch, comprising the following steps:

[0008] Step S1, establishing a feature extraction unit for extracting features of active ingredients in the sleep patch and extracting feature information thereof;

[0009] Step S2, establishing a feature value comparison unit, which includes a feature value storage module and a feature value comparison module, wherein the feature value storage module stores various feature information of different categories of active ingredients and the corresponding proportions of various feature information, compares the feature information extracted by the feature extraction unit with the feature information of each active ingredient, and obtains the corresponding active ingredient;

[0010] Step S3, establishing a characteristic value statistical unit for counting the proportions corresponding to the characteristic information of various effective components in the same sample, and storing the statistical results in a database;

[0011] Step S4, establishing an effective component analysis unit for conducting safety analysis on the effective components in the sample and reasonably allocating effective components in different proportions.

[0012] Preferably, in step S2, the specific process of establishing the feature extraction unit is:

[0013] 2.1, based on the gas chromatography-mass spectrometer GC-MS for detection, the instrument consists of two parts: gas chromatograph GC and mass spectrometer MS; the gas chromatograph separates various mixtures by programmed temperature rise, and different substances have different retention times in the chromatographic column. The substances are separated according to the order of peaks, and their relative contents are obtained through the detector; the mass spectrometer is used to provide material structure information, that is, the substances separated by the gas chromatograph GC are introduced into the mass spectrometer to obtain the mass spectrum, and the molecular formula and structural formula are obtained through the mass spectrum;

[0014] 2.2, the separated active ingredient mixed liquid is placed in a gas chromatograph GC for detection, the peak time is determined by the gas chromatograph GC, and the peak time corresponding to each substance is collected, the molecular formula structure is determined by a mass spectrometer, and the molecular formula and relative content of each substance in the mixed liquid are obtained;

[0015] 2.3, extract characteristic information, including peak time, molecular formula, relative content, and enter it into the database.

[0016] Preferably, the specific steps of step S2 are:

[0017] 3.1, comparing the feature information extracted by the input feature extraction unit with the feature information of the active ingredient in the database;

[0018] 3.2, the database stores characteristic information of various active ingredients and the proportion information of each characteristic information, including the peak time of various active ingredients , molecular formula , relative content

[0019] 3.3, set the characteristic information threshold of the active ingredient. If it exists in the above databases, calculate the proportion of each characteristic information by formula (1):

[0020] (1) ;

[0021] in, Indicates The active ingredient The proportion of active ingredients, Indicates The active ingredient The peak time of the active ingredients, Indicates The active ingredient The molecular formula of the active ingredient, Indicates The active ingredient The relative content of active ingredients, It represents the sum of the products of the characteristic information of all effective ingredients;

[0022] 3.4. If any feature information in the above database exceeds the threshold, it is considered that no useful features can be extracted, the test is terminated, and the sample feature information and active ingredient classification category are re-entered;

[0023] 3.5. If the characteristic information does not exist in the above database, select an active ingredient category from the active ingredients to be detected As active ingredients, , use the feature value comparison algorithm to identify whether the active ingredient contains the target active ingredient, and set the category of the active ingredient to be detected as , the effective ingredient is analyzed by formula (2) Whether there is an active ingredient of the same category in the database :

[0024] (2);

[0025] in, Indicates The active ingredient The peak time of the active ingredients, Indicates The active ingredient The peak time of the active ingredients, For the The characteristic value threshold of the peak time of the active ingredient;

[0026] 3.6, if formula (2) is not 0 and is less than or equal to , it indicates that the active ingredient of the test Contains active ingredients , the difference between the two molecular formulas is calculated by formula (3):

[0027] (3);

[0028] in, Indicates The molecular formula of the active ingredient, Indicates The molecular formula of the active ingredient;

[0029] 3.7, if formula (3) is not 0, the effective component is calculated by formula (4) , The proportion of active ingredients corresponding to the largest proportion of active ingredients:

[0030] (4);

[0031] in, and They represent the Active ingredients and The effective ingredient ratios of the three active ingredients were calculated by formula (1);

[0032] 3.8, record the comparison results and the active ingredients to be analyzed As active ingredient and increase the corresponding active ingredient classification category by 1;

[0033] 3.9, if formula (2) is greater than , then all the classifications are compared by formula (5) to obtain the effective components to be analyzed. Active ingredients belonging to the same general category:

[0034] (5);

[0035] in, Indicates The peak time of the active ingredients, Indicates The peak time of the active ingredients, is the characteristic value threshold of the peak time;

[0036] 3.10, select the active ingredient i to be analyzed and compare it with the active ingredient j obtained by analysis, and calculate the active ingredient ratio corresponding to active ingredients i and j by formula (6):

[0037] (6);

[0038] in, and Respectively represent the effective components calculated by formula (1) , The corresponding proportion of active ingredients;

[0039] 3.11, record the comparison results and the active ingredients to be analyzed As active ingredient and increase the classification categories of active ingredients;

[0040] 3.12, determine whether the analysis of all active ingredients in the sample to be tested has been completed, if not completed, repeat step S1;

[0041] 3.13, if completed, the final effective ingredient ratio is calculated by formula (1) to obtain various effective ingredients.

[0042] Preferably, the specific steps of step S4 are:

[0043] 4.1, calculate the proportion of each active ingredient by formula (1);

[0044] 4.2. Use safety assessment algorithms to screen and rate the active ingredients that meet safety standards;

[0045] 4.3. Rationally distribute the active ingredients based on safety ratings.

[0046] Preferably, the specific steps of the feature extraction unit for extracting features from samples are: first, setting time parameters, collecting feature information of effective components in each time period, and forming a feature information matrix; extracting feature information of effective components in each time period and their proportions, and forming feature information eigenvalues; wherein the proportion corresponding to each eigenvalue is calculated by formula (7):

[0047] (7);

[0048] in, Indicates The proportion of active ingredients, Indicates The peak time of the active ingredients, Indicates The molecular formula of the active ingredient, Indicates The relative content of active ingredients, Represents the sum of the products of the characteristic information of all effective ingredients.

[0049] Preferably, the eigenvalue storage module includes the peak emergence time, molecular formula, and relative content, and the eigenvalue comparison module comprehensively evaluates the comparison results of each eigenvalue to obtain a comprehensive comparison result.

[0050] Preferably, the specific method for the safety assessment in step S4 is as follows: First, set safety standards, set standard ratios for each active ingredient, including an upper limit ratio and a lower limit ratio. Among them, the upper limit ratio means that the total ratio of this ingredient shall not exceed this ratio, and the lower limit ratio means that the total ratio of this ingredient shall not be lower than this ratio; calculate the ratio of each active ingredient through formula (1). If it exceeds the upper limit ratio, adjust the safety assessment level of this active ingredient to the second level. If it is lower than the lower limit ratio, adjust the safety assessment level of this active ingredient to the fourth level, and the rest are the third level; then, sort the active ingredients according to their corresponding safety assessment levels. If there is no safety level higher than the third level, it is considered that the safety assessment is qualified.

[0051] Preferably, if some active ingredients belong to the second level of safety assessment, set the number of ingredients with the second level of safety assessment as A, and the number of ingredients with the third level of safety assessment as B. If A≥B, randomly allocate the A ingredients with the second level of safety assessment. If A<B, randomly select A + 1 ingredients with the second level of safety assessment, and evenly distribute the remaining B - A ingredients with the third level of safety assessment according to the ratio until A = B; if A<B, randomly select A ingredients with the second level of safety assessment, delete the ingredient with the second largest ratio among the remaining B - A + 1 ingredients with the third level of safety assessment, and evenly distribute the other ingredients according to the ratio until A = B.

[0052] Preferably, construct an active ingredient model according to the safety assessment results, and combine the detection data of the sleep patch to establish a database of the active ingredients of the sleep patch, which is convenient for producers to search the corresponding database in time, adjust the production process, reduce the addition ratio of active ingredients, and improve the utilization rate of active ingredients.

[0053] Preferably, the feature extraction unit uses a gas chromatography-mass spectrometer GC-MS to detect the effective ingredients. The feature extraction unit also includes a data preprocessing module for preprocessing the sample, collecting the characteristic information and the proportion of the effective ingredients in each time period, and forming a characteristic value of the characteristic information. The safety assessment algorithm is a voting algorithm. The safety rating in the above safety assessment is used to set different weights corresponding to different safety levels, and the number of votes for each effective ingredient is counted in turn to obtain the final safety assessment result. When the reasonable allocation in step S4 is performed, the effective ingredients are randomly allocated in a 1:1 ratio according to the average proportion of each effective ingredient. The time parameters are 3-6 minutes, 6-13 minutes and more than 13 minutes. The effective ingredients are extracted in each time period to form a characteristic value of the characteristic information. The proportion corresponding to each characteristic value is calculated by formula (7).

[0054] The present invention has the following beneficial effects:

[0055] First, the present invention significantly improves the recognition rate of active ingredients. By establishing a multi-dimensional feature extraction and comparison mechanism, the present method can accurately distinguish compounds with similar structures, greatly reducing misjudgments and omissions. In particular, for some ingredients with low content but important physiological activities, the present method shows excellent detection capabilities. This not only helps to comprehensively control product quality, but also provides the possibility for the discovery of new active ingredients.

[0056] Secondly, the present invention has made significant progress in terms of quantitative accuracy. By comprehensively considering multiple indicators such as peak time, mass spectrum and relative content, and introducing an innovative mathematical model, this method effectively overcomes the limitations of the traditional peak area normalization method and greatly improves the reliability of the quantitative results. This is of great significance for ensuring the consistency and stability of product formulas.

[0057] Furthermore, the present invention greatly improves the detection efficiency. With the help of an intelligent data processing system, the present method shortens the analysis time to less than half of the traditional method. This efficiency improvement can not only meet the needs of large-scale production, but also provide technical support for real-time quality control in the production process, which helps to improve the overall production efficiency.

[0058] More importantly, the present invention has achieved a qualitative leap in safety assessment. By establishing a comprehensive safety assessment system, the present method not only considers the content of a single ingredient, but also simulates the combined effects that may be produced when multiple ingredients coexist. This comprehensive assessment method greatly improves the accuracy and reliability of safety assessment, providing consumers with a stronger safety guarantee.

[0059] In addition, the method of the present invention has good adaptability and scalability. Although initially developed for sleep patches, the method can also be applied to other types of transdermal drug delivery systems through appropriate adjustment and optimization. This wide applicability lays the foundation for the promotion and application of the present invention in related fields.

[0060] In general, the intelligent detection method for the active ingredients of sleep patches proposed in the present invention not only solves many problems in the prior art by comprehensively improving the accuracy, efficiency and safety assessment level of the detection, but also provides strong technical support for the quality control and innovative development of sleep patch products. Today, when people are paying more and more attention to sleep health, the present invention will undoubtedly make important contributions to improving the quality of sleep patch products and promoting the healthy development of the industry. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 It is a logic block diagram of the method of the present invention.

[0062] Figure 2 It is a logic block diagram of the feature extraction unit of the present invention.

[0063] Figure 3 It is a logic block diagram of the characteristic value comparison unit of the present invention.

[0064] Figure 4 It is a logic block diagram of the characteristic value statistical unit of the present invention.

[0065] Figure 5 It is a logic block diagram of the active ingredient analysis unit of the present invention. DETAILED DESCRIPTION

[0066] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the specific implementation methods, structures, features and effects thereof are described in detail below in conjunction with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0067] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0068] See also Figure 1-5 The present invention discloses an intelligent detection method for the active ingredients of a sleeping patch, which aims to improve the detection efficiency and accuracy of the active ingredients of the sleeping patch and provide valuable data support for manufacturers.

[0069] First, a method for intelligently detecting active ingredients of a sleep patch of the present invention comprises the following steps: establishing a feature extraction unit 10, a feature value comparison unit 20, a feature value statistics unit 30 and an active ingredient analysis unit 40. These units work together to complete the intelligent detection process of the active ingredients of the sleep patch.

[0070] Specifically, the feature extraction unit 10 is responsible for extracting the feature information of the active ingredients from the sleep patch sample. These feature information usually include key data such as peak time, molecular formula and relative content. The feature value comparison unit 20 includes a feature value storage module 21 and a feature value comparison module 22. The feature value storage module 21 pre-stores the feature information of various known active ingredients and their corresponding proportions, while the feature value comparison module 22 is responsible for comparing the extracted feature information with the known data to identify the active ingredients in the sample.

[0071] Next, the characteristic value statistics unit 30 will count the characteristic information proportions of various active ingredients in the same sample and store the results in the database. Finally, the active ingredient analysis unit 40 is responsible for safety analysis of the active ingredients in the sample and reasonably distributes the active ingredients according to different proportions.

[0072] In one embodiment of the present invention, the establishment process of the feature extraction unit 10 is as follows: First, a gas chromatography-mass spectrometer (GC-MS) is used for detection. The instrument consists of a gas chromatograph (GC) and a mass spectrometer (MS). The gas chromatograph separates the mixture by programming the temperature. Different substances have different retention times in the chromatographic column, so that the substances are separated according to the order of peaks, and their relative contents are obtained through the detector. The mass spectrometer is used to provide structural information of the substance. It imports the substance separated by the gas chromatograph to obtain a mass spectrum, and then derives its molecular formula and structural formula.

[0073] Then, the separated active ingredient mixture is placed in a gas chromatograph for detection, the peak time is determined by the gas chromatograph, and the peak time corresponding to each substance is collected. At the same time, the molecular formula structure is determined by a mass spectrometer to obtain the molecular formula and relative content of each substance in the mixture.

[0074] Finally, feature information is extracted, including peak time, molecular formula, and relative content, and this information is input into the database. In this way, the feature extraction unit completes the comprehensive extraction of the features of the active ingredients in the sleep patch sample.

[0075] Preferably, the specific implementation process of step S2 is as follows: First, the feature information extracted by the feature extraction unit is compared with the feature information of known active ingredients in the database. The database stores the feature information of various active ingredients and their specific gravity information, including the peak time of various active ingredients. , molecular formula , relative content .

[0076] Secondly, set the characteristic information threshold of the active ingredient. If the corresponding characteristic information exists in the database, the proportion of each characteristic information is calculated by formula (1):

[0077] (1);

[0078] in, Indicates The active ingredient The proportion of active ingredients, Indicates The active ingredient The peak time of the active ingredients, Indicates The active ingredient The molecular formula of the active ingredient, Indicates The active ingredient The relative content of active ingredients, Represents the sum of the products of the characteristic information of all effective ingredients.

[0079] Again, if any feature information in the database exceeds the preset threshold, it is considered that no useful features can be extracted. At this time, the current test will be terminated and the sample feature information and active ingredient classification category will be required to be re-entered. This preset threshold is usually determined based on experiments or experience. For example, the threshold of peak time can be set to ±0.5 minutes, and the threshold of molecular weight can be set to ±5amu.

[0080] If the characteristic information does not exist in the database, select a category from the active ingredient to be detected ( ) as the active ingredient, and use the feature value comparison algorithm to identify whether the active ingredient contains the target active ingredient. Set the category of the active ingredient to be detected as i, and use formula (2) to analyze whether the active ingredient i and the database have the same category of active ingredients to be analyzed :

[0081] (2);

[0082] in, Indicates The active ingredient The peak time of the active ingredients, Indicates The active ingredient The peak time of the active ingredients, For the The characteristic value threshold of the peak time of the active ingredient. This threshold ε is usually set to 0.1 minutes, because in actual detection, the peak time fluctuation of the same substance usually does not exceed this range.

[0083] If formula (2) is not 0 and is less than or equal to , it indicates that the active ingredient i detected contains active ingredients Then, the difference between the two molecular formulas is calculated using formula (3):

[0084] (3);

[0085] Wherein, mix represents the molecular formula of the i-th active ingredient, Represents the molecular formula of the jth active ingredient.

[0086] If formula (3) is not 0, the effective component is calculated by formula (4) , The proportion of active ingredients corresponding to the largest proportion of active ingredients:

[0087] (4);

[0088] in, and They represent the Active ingredients and The effective ingredient ratios of the three active ingredients were calculated by formula (1).

[0089] Then, record the comparison results and take the active ingredient i to be analyzed as the active ingredient , and increase the corresponding active ingredient classification category by 1.

[0090] If formula (2) is greater than , then all the categories are compared by formula (5) to obtain the effective ingredients belonging to the same category as the effective ingredient i to be analyzed:

[0091] (5);

[0092] in, Indicates The peak time of the active ingredients, Indicates The peak time of the active ingredients, The characteristic value threshold of the peak time is usually set to 0.5 minutes, because the retention time difference of the same type of active ingredients in the chromatographic column usually does not exceed this range.

[0093] Next, the active ingredient i to be analyzed is selected and compared with the active ingredient j with the largest proportion among the analyzed active ingredients, and the proportion of the active ingredients corresponding to the active ingredients i and j is calculated by formula (6):

[0094] (6);

[0095] in, and Respectively represent the effective components calculated by formula (1) , The corresponding proportion of active ingredients.

[0096] Finally, record the comparison results, take the active ingredient i to be analyzed as active ingredient j, and increase the active ingredient classification category by 1. Repeat the above steps until the analysis of all active ingredients in the sample to be tested is completed. After completion, calculate the final active ingredient ratio through formula (1) to obtain each type of active ingredient.

[0097] In another embodiment of the present invention, the specific implementation process of step S4 is as follows: First, the proportion of each active ingredient is calculated by formula (1). Then, the safety assessment algorithm is used to screen and the active ingredients that meet the safety standards are rated for safety. Finally, the active ingredients are reasonably allocated according to the safety rating.

[0098] Preferably, the specific steps of the feature extraction unit for extracting features from samples are: first, set time parameters, collect feature information of active ingredients in each time period, and form a feature information matrix. Then, extract the feature information and its proportion of active ingredients in each time period to form feature information eigenvalues. The proportion corresponding to each eigenvalue is calculated by the following formula:

[0099] (7);

[0100] in, Indicates The proportion of active ingredients Indicates The peak time of the active ingredients, Indicates The molecular formula of the active ingredient, Indicates The relative content of active ingredients, Represents the sum of the products of the characteristic information of all effective ingredients.

[0101] In the present invention, the eigenvalue storage module 21 includes information such as peak emergence time, molecular formula, relative content, etc., while the eigenvalue comparison module 22 comprehensively evaluates the comparison results of various eigenvalues to obtain a comprehensive comparison result. This comprehensive evaluation can more comprehensively reflect the characteristics of the active ingredients and improve the accuracy of detection.

[0102] In terms of safety assessment, the present invention adopts an innovative method. First, set standard ratios for each active ingredient, including an upper limit ratio and a lower limit ratio. The upper limit ratio represents the maximum value that the total ratio of this ingredient shall not exceed, and the lower limit ratio represents the minimum value that the total ratio of this ingredient shall not be lower than. For example, for a certain common calming ingredient, the upper limit ratio can be set at 20% and the lower limit ratio at 5%.

[0103] Then, calculate the ratio of each active ingredient through formula (1). If the ratio of a certain active ingredient exceeds the upper limit, adjust its safety assessment level to the second level; if it is lower than the lower limit, adjust it to the fourth level; the rest are at the third level. This grading method can effectively identify potential safety risks and provide a basis for subsequent ingredient adjustment.

[0104] Finally, sort the active ingredients according to their corresponding safety assessment levels. If there are no ingredients with a safety level higher than the third level, it is considered that this batch of sleep patches passes the safety assessment.

[0105] In a preferred embodiment of the present invention, if some active ingredients belong to the second level of safety assessment, the following strategy is adopted for adjustment: set the number of ingredients at the second level of safety assessment as A, and the number of ingredients at the third level as B. If A≥B, randomly allocate the A ingredients at the second level of safety assessment. If A<B, randomly select A + 1 ingredients at the second level of safety assessment, and evenly distribute the remaining B - A ingredients at the third level of safety assessment according to the ratio until A = B. This method can, while ensuring safety, maximize the retention of the diversity of active ingredients.

[0106] In addition, the present invention also proposes a method for constructing an active ingredient model based on the safety assessment results. By combining the detection data of the sleep patches, establish a database of the active ingredients of the sleep patches, which is convenient for producers to search the corresponding database in a timely manner, adjust the production process, reduce the addition ratio of active ingredients, and improve the utilization rate of active ingredients. This can not only improve the product quality but also reduce the production cost, achieving a win-win situation in economic and social benefits.

[0107] Finally, the feature extraction unit of the present invention uses a gas chromatography-mass spectrometer (GC-MS) to detect the active ingredients, and also includes a data preprocessing module for preprocessing the sample, collecting the feature information and its proportion of the active ingredients in each time period, and forming a feature value of the feature information. This preprocessing can effectively improve the accuracy and efficiency of subsequent analysis.

[0108] The safety assessment algorithm used in the present invention is a voting algorithm. Using the safety rating in the above safety assessment, different weights are set for different safety levels, and the votes for each active ingredient are counted in turn to obtain the final safety assessment result. For example, the weight of the second level can be set to 3, the weight of the third level can be set to 2, and the weight of the fourth level can be set to 1. This weighted voting method can more accurately reflect the comprehensive safety of each active ingredient.

[0109] In terms of the rational distribution of active ingredients, the present invention adopts an innovative random distribution method. Specifically, the active ingredients are randomly distributed in a 1:1 ratio according to the average proportion of each active ingredient. This method can minimize the risk of excessive concentration of a single ingredient while ensuring effectiveness.

[0110] It is worth noting that in the feature extraction process, the present invention uses three time periods for analysis: 3-6 minutes, 6-13 minutes, and more than 13 minutes. This segmentation method is based on a large amount of experimental data and can effectively cover the peak time range of most active ingredients in sleep patches. In each time period, the active ingredients are feature extracted to form feature information eigenvalues. The proportions corresponding to these eigenvalues ​​are calculated by formula (7), thereby obtaining a comprehensive and detailed distribution map of the active ingredients.

[0111] In general, the intelligent detection method for active ingredients of sleep patches proposed in the present invention realizes rapid, accurate and comprehensive detection of active ingredients of sleep patches by combining gas chromatography-mass spectrometry analysis technology, data mining algorithm and safety assessment system. This method not only improves detection efficiency, but also provides valuable data support for manufacturers, which helps to improve product formula and production process.

[0112] A significant advantage of the present invention is its flexibility and adaptability. By adjusting the parameter settings of the feature extraction unit 10, the method can be easily adapted to different types of sleep patch products. For example, for sleep patches containing Chinese herbal ingredients, the temperature rise program of the gas chromatography can be appropriately extended to better separate complex Chinese herbal mixtures. At the same time, the scanning range of the mass spectrometer can also be optimized according to the molecular weight of the target component, thereby improving the sensitivity and accuracy of the detection.

[0113] Another feature worth noting is the feature value comparison unit 20 of the present invention. This unit can not only identify known active ingredients, but also discover potential new ingredients. This is achieved through an innovative algorithm: when a detected ingredient does not completely match a known ingredient in the database, the system will automatically mark it as a component to be studied and record its feature information. This mechanism provides the possibility of discovering new sleep-aiding ingredients and has important research value.

[0114] In practical applications, the method of the present invention also demonstrates excellent anti-interference ability. Sleep patches usually contain a variety of excipients, such as adhesives, preservatives, etc., which may interfere with the detection of active ingredients. The present invention effectively reduces the impact of these interfering factors by optimizing the sample pretreatment method and setting a reasonable feature information threshold. For example, in the sample pretreatment stage, a specific solvent system can be used to selectively extract the target component while removing interfering substances to the greatest extent.

[0115] The safety assessment system of the present invention also has unique advantages. Traditional safety assessment methods often only focus on whether the content of a single component exceeds the standard, while ignoring the interaction between components. The multi-level assessment system proposed by the present invention not only takes into account the content of a single component, but also simulates the synergistic or antagonistic effects that may occur when multiple components coexist through complex algorithms. This comprehensive assessment method greatly improves the accuracy and reliability of safety assessments.

[0116] In terms of data management, the active ingredient database of the sleep patch established by the present invention has the ability of self-learning and self-updating. The results of each test are recorded and used to update the database, which enables the system to continuously optimize its detection and evaluation algorithms. As the number of test samples increases, the accuracy and reliability of the system will continue to improve. This dynamic update mechanism enables the present invention to always maintain its technological advancement.

[0117] The present invention also pays special attention to the interpretability of the test results. When outputting the test report, the system not only gives the content and safety rating of various active ingredients, but also provides a detailed analysis process and reasoning basis. This transparent reporting mechanism helps producers and regulators better understand the test results and make more reasonable decisions.

[0118] In actual production, the method of the present invention can also be seamlessly connected with the intelligent manufacturing system. By real-time monitoring of product quality on the production line, the system can promptly detect and correct deviations in the production process, thereby ensuring the stability and consistency of product quality. This not only improves production efficiency, but also greatly reduces the rate of defective products.

[0119] It is worth mentioning that the method of the present invention also has good scalability. Although it is currently mainly used for the detection of sleep patches, through appropriate adjustment and optimization, the method can also be applied to other types of transdermal drug delivery systems, such as pain relief patches, hormone patches, etc. This wide applicability greatly improves the commercial value and social value of the present invention.

[0120] Finally, the present invention also takes environmental friendliness into special consideration. During sample pretreatment and detection, low-toxic and easily degradable reagents and solvents are selected as much as possible to minimize environmental pollution. At the same time, by optimizing the detection process, energy consumption and waste generation are also greatly reduced. This green and sustainable design concept not only meets current environmental protection requirements, but also creates a good social image for the company.

[0121] In general, the intelligent detection method for the active ingredients of the sleep patch proposed in the present invention realizes the rapid, accurate and comprehensive detection of the active ingredients of the sleep patch by comprehensively using advanced analytical techniques, innovative algorithm design and a comprehensive safety assessment system. This method not only improves the efficiency and accuracy of product quality control, but also provides strong technical support for product formula optimization and the research and development of new ingredients. In the current context of increasing attention to sleep health, the present invention will undoubtedly make important contributions to the quality improvement and market development of sleep patch products.

[0122] In order to verify the superiority of the intelligent detection method for active ingredients of the sleep patch proposed in the present invention, we conducted a series of comparative experiments. The specific operation process, detection results and analysis of the embodiments and comparative examples will be described in detail below.

[0123] Example 1: We selected a common lavender essential oil sleeping patch on the market as the test object. First, a feature extraction unit was established according to the method of the present invention, and gas chromatography-mass spectrometry (GC-MS) was used for detection. The chromatographic column selected was a DB-5MS capillary column (30m×0.25mm×0.25μm), and the carrier gas was high-purity helium. The heating program was: initial temperature 60°C, maintained for 2min, increased to 250°C at 5°C / min, and maintained for 5min. The mass spectrometer used an electron impact (EI) ion source, the ion source temperature was 230°C, the interface temperature was 280°C, and the scanning range was m / z 35-500.

[0124] The sample pretreatment adopts ultrasonic extraction method, taking 1g of sleep patch sample, adding 10mL of ethanol, ultrasonic extraction for 30min, and taking the supernatant for GC-MS analysis after centrifugation. We process and analyze the test data according to the characteristic value comparison unit, characteristic value statistics unit and effective component analysis unit established by the method of the present invention.

[0125] Comparative Example 1: Traditional qualitative and quantitative analysis methods were used. The same GC-MS instrument and chromatographic conditions were used, but the data processing adopted the conventional peak area normalization method for quantification, and the qualitative analysis was performed by comparing with the standard spectral library. The safety assessment only considered whether a single component exceeded the standard.

[0126] Testing indicators and methods:

[0127] 1. Active ingredient recognition rate: Calculate the percentage of correctly identified active ingredients to the total number of active ingredients.

[0128] 2. Quantitative accuracy: percentage deviation from a standard of known concentration.

[0129] 3. Detection sensitivity: minimum detectable concentration (LOD).

[0130] 4. Analysis time: the time required from sample preparation to issuing a complete report.

[0131] 5. Safety assessment accuracy: the percentage of consistency with expert assessment results.

[0132] The test results are shown in the following table:

[0133] Detection indicators Example 1 Comparative Example 1 Active ingredient recognition rate (%) 98.5% 85.3% Quantitative accuracy (%) 96.8% 91.2% Detection sensitivity (LOD) 0.01mg / kg 0.1mg / kg Analysis time 45 minutes 120 minutes Safety assessment accuracy (%) 97.6% 82.4%

[0134] From the above results, it can be seen that the method of the present invention is significantly superior to the traditional method in all indicators. Let us analyze the significance of these results in depth:

[0135] 1. Recognition rate of active ingredients: The recognition rate of the method of the present invention reached 98.5%, which is much higher than the 85.3% of the traditional method. This shows that our feature extraction unit and feature value comparison unit have stronger recognition ability, especially for compounds with similar structures. For example, in lavender essential oil, the structures of linalool and linalyl acetate are very similar. The traditional method easily confuses them, while the method of the present invention successfully distinguishes the two compounds by comprehensively considering the peak time, mass spectrum and relative content.

[0136] 2. Quantitative accuracy: The quantitative accuracy of the method of the present invention reaches 96.8%, which is 5.6 percentage points higher than the traditional method. This is due to our characteristic value statistical unit, which not only considers the peak area, but also incorporates the stability of the peak time into the calculation, effectively reducing the influence of factors such as baseline drift.

[0137] 3. Detection sensitivity: The minimum detectable concentration (LOD) of the method of the present invention reaches 0.01 mg / kg, which is one order of magnitude lower than that of the traditional method. This is mainly due to our optimized sample pretreatment method and data processing algorithm. For example, for α-pinene (α-Pinene), which has extremely low content but important physiological activity, the traditional method may ignore it, but the method of the present invention can accurately detect and quantify it.

[0138] 4. Analysis time: The method of the present invention shortens the analysis time to 45 minutes, which is less than half of the traditional method. This efficiency improvement mainly comes from our intelligent data processing system. Especially in the analysis of complex samples, the traditional method may require manual confirmation of each chromatographic peak one by one, while the present invention has a higher degree of automation, which greatly saves time.

[0139] 5. Safety assessment accuracy: The safety assessment accuracy of the method of the present invention is as high as 97.6%, far exceeding the 82.4% of the traditional method. This advantage comes from our comprehensive safety assessment system. Traditional methods often only consider whether a single ingredient exceeds the standard, but ignore the interaction between ingredients. For example, although the content of certain terpenoid compounds is not high alone, they may produce a synergistic effect and enhance the sedative effect when they exist together. Our method simulates the combined effects of multiple ingredients to provide a more accurate safety assessment.

[0140] Based on the above results, we believe that Example 1 is the best example of the present invention. In this example, we give full play to the advantages of each unit of the present invention and achieve rapid, accurate and comprehensive detection of the active ingredients of the sleep patch.

[0141] In general, these test results clearly demonstrate the significant advantages of the method of the present invention in the field of intelligent detection of active ingredients in sleep patches. It not only improves the accuracy and efficiency of detection, but also provides more reliable technical support for product quality control and safety assessment. In the current context where people are paying more and more attention to sleep quality, this invention will undoubtedly make an important contribution to the quality improvement and market development of sleep patch products.

[0142] It should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. An intelligent detection method for active ingredients of a sleeping patch, characterized in that: The steps include: Step S1, establish a feature extraction unit for extracting features of the active ingredients in the sleep patch and extracting its feature information ； Step S2, establishing a feature value comparison unit, which includes a feature value storage module and a feature value comparison module, wherein the feature value storage module stores various feature information of different categories of active ingredients and the corresponding proportions of various feature information, compares the feature information extracted by the feature extraction unit with the feature information of each active ingredient, and obtains the corresponding active ingredient; Step S3, establishing a characteristic value statistical unit for counting the proportions corresponding to the characteristic information of various effective components in the same sample, and storing the statistical results in a database; Step S4, establishing an effective component analysis unit for conducting safety analysis on the effective components in the sample and reasonably distributing effective components in different proportions; The specific steps of step S2 are: S3.1, comparing the feature information extracted by the input feature extraction unit with the feature information of the active ingredient in the database; S3.2, the database stores characteristic information of various active ingredients and the proportion information of each characteristic information, including the peak time of various active ingredients , molecular formula , relative content ; S3.3, set the characteristic information threshold of the active ingredient. If it exists in all the above databases, calculate the proportion of each characteristic information by formula (1): (1); in, Indicates The active ingredient The proportion of active ingredients, Indicates The active ingredient The peak time of the active ingredients, Indicates The active ingredient The molecular formula of the active ingredient, Indicates The active ingredient The relative content of active ingredients, It represents the sum of the products of the characteristic information of all effective ingredients; S3.4, if any feature information in the above database exceeds the threshold, it is considered that no useful features can be extracted, the test is terminated, and the sample feature information and active ingredient classification category are re-entered; S3.5, if the characteristic information does not exist in the above database, select an active ingredient category from the active ingredients to be detected As active ingredients, , use the feature value comparison algorithm to identify whether the active ingredient contains the target active ingredient, and set the category of the active ingredient to be detected as , the effective ingredient is analyzed by formula (2) Whether there is an active ingredient of the same category in the database (2); in, Indicates The active ingredient The peak time of the active ingredients, Indicates The active ingredient The peak time of the active ingredients, For the The characteristic value threshold of the peak time of the active ingredient; S3.6, if formula (2) is not 0, and is less than or equal to This indicates that the active ingredient Contains active ingredients , the difference between the two molecular formulas is calculated by formula (3): (3); in, Indicates The molecular formula of the active ingredient, Indicates The molecular formula of the active ingredient; S3.7, if formula (3) is not 0, the active ingredient is calculated by formula (4) , The proportion of active ingredients corresponding to the largest proportion of active ingredients: (4); in, and They represent the Active ingredients and The effective ingredient ratios of the three active ingredients were calculated by formula (1); S3.8, record the comparison results and the active ingredients to be analyzed As active ingredient and increase the corresponding active ingredient classification category by 1; S3.9, if formula (2) is greater than , then all the classifications are compared by formula (5) to obtain the effective components to be analyzed. Active ingredients belonging to the same general category: (5); in, Indicates The peak time of the active ingredients, Indicates The peak time of the active ingredients, is the characteristic value threshold of the peak time; S3.10, select the active ingredient i to be analyzed and compare it with the active ingredient j obtained by analysis with the largest active ingredient ratio, and calculate the active ingredient ratio corresponding to the active ingredients i and j by formula (6): (6); in, and Respectively represent the effective components calculated by formula (1) , The corresponding proportion of active ingredients; S3.11, record the comparison results and the active ingredients to be analyzed As active ingredient and increase the classification categories of active ingredients; S3.12, determine whether the analysis of all active ingredients in the sample to be tested has been completed, if not completed, repeat step S1; S3.13, if completed, calculate the final effective ingredient ratio using formula (1) to obtain each type of effective ingredient.

2. A method for intelligently detecting active ingredients of a sleeping patch as claimed in claim 1, characterized in that: In step S2, the specific process of establishing the feature extraction unit is as follows: S2.1, detection based on gas chromatography-mass spectrometry (GC-MS), the gas chromatography-mass spectrometry (GC-MS) comprising two parts: a gas chromatograph (GC) and a mass spectrometer (MS); wherein the gas chromatograph separates various mixtures by programmed temperature rise, and different substances have different retention times in the chromatographic column. The substances are separated according to the order of peaks, and their relative contents are obtained through a detector; the mass spectrometer is used to provide information on the structure of the substance, that is, the substance separated by the gas chromatograph (GC) is introduced into the mass spectrometer to obtain a mass spectrum, and its molecular formula and structural formula are obtained through the mass spectrum; S2.2, placing the separated active ingredient mixed liquid into a gas chromatograph GC for detection, determining the peak time by the gas chromatograph GC, collecting the peak time corresponding to each substance, determining the molecular formula structure by a mass spectrometer, and obtaining the molecular formula and relative content of each substance in the mixed liquid; S2.3, extract characteristic information, including peak time, molecular formula, relative content, and enter into the database.

3. A method for intelligently detecting active ingredients of a sleeping patch as claimed in claim 1, characterized in that: The specific steps of step S4 are: S4.1, calculate the proportion of each active ingredient by formula (1); S4.2, use the safety assessment algorithm to screen and rate the safety of active ingredients that meet the safety standards; S4.3, distribute the active ingredients rationally according to the safety rating.

4. A method for intelligently detecting active ingredients of a sleeping patch as claimed in claim 3, characterized in that: The specific steps of the feature extraction unit for extracting features from samples are as follows: first, set the time parameters, collect the feature information of the effective components in each time period, and form a feature information matrix; extract the feature information and its proportion of the effective components in each time period, and form a feature information eigenvalue; wherein the proportion corresponding to each eigenvalue is calculated by formula (7): (7); in, Indicates The proportion of active ingredients, Indicates The peak time of the active ingredients, Indicates The molecular formula of the active ingredient, Indicates The relative content of active ingredients, Represents the sum of the products of the characteristic information of all effective ingredients.

5. A method for intelligently detecting active ingredients of a sleeping patch as claimed in claim 4, characterized in that: The characteristic value storage module includes peak time, molecular formula, and relative content, and the characteristic value comparison module comprehensively evaluates the comparison results of various characteristic values ​​to obtain a comprehensive comparison result.

6. A method for intelligently detecting active ingredients of a sleeping patch as claimed in claim 5, characterized in that: The specific method of the safety assessment in step S4 is as follows: first, set a safety standard, set a standard proportion for each active ingredient, including an upper limit proportion and a lower limit proportion, wherein the upper limit proportion indicates that the total proportion of the ingredient shall not exceed the proportion, and the lower limit proportion indicates that the total proportion of the ingredient shall not be lower than the proportion; calculate the proportion of each active ingredient by formula (1); if it exceeds the upper limit proportion, the safety assessment level of the active ingredient is adjusted to the second level; if it is lower than the lower limit proportion, the safety assessment level of the active ingredient is adjusted to the fourth level, and the rest are the third level; then, sort the active ingredients according to the corresponding safety assessment levels, and if there is no active ingredient with a safety level higher than the third level, the safety assessment is considered qualified.

7. A method for intelligently detecting active ingredients of a sleeping patch as claimed in claim 6, characterized in that: If some active ingredients belong to the second level of safety assessment, the number of ingredients in the second level of safety assessment is set to A, and the number of ingredients in the third level is set to B. If A ≥ B, A second-level ingredients are randomly assigned; if A <B: a. Randomly select A second-level components; b. Evenly distribute the remaining BA third-level components according to their proportions until A=B; c. If a balance cannot be achieved in step b, remove the second largest component in the third level and redistribute the remaining components until A=B.

8. A method for intelligently detecting active ingredients of a sleeping patch as claimed in claim 7, characterized in that: An active ingredient model is constructed based on the safety assessment results, and a database of active ingredients for sleeping patches is established in combination with the test data of sleeping patches. This allows manufacturers to find the corresponding database in a timely manner, adjust the production process, reduce the proportion of active ingredients added, and improve the utilization rate of active ingredients.

9. A method for intelligently detecting active ingredients of a sleeping patch as claimed in claim 8, characterized in that: The feature extraction unit uses a gas chromatography-mass spectrometer GC-MS to detect the effective ingredients. The feature extraction unit also includes a data preprocessing module for preprocessing the sample, collecting the characteristic information and the proportion of the effective ingredients in each time period, and forming a characteristic value of the characteristic information. The safety assessment algorithm is a voting algorithm. The safety rating in the above safety assessment is used to set different weights corresponding to different safety levels, and the number of votes for each effective ingredient is counted in turn to obtain a final safety assessment result. When the reasonable allocation in step S4 is specifically performed according to the average value of the proportion of each effective ingredient, the effective ingredients are randomly allocated in a 1:1 ratio; the time parameters are 3-6 minutes, 6-13 minutes and more than 13 minutes, and the effective ingredients are feature extracted in each time period to form a characteristic value of the characteristic information; wherein the proportion corresponding to each characteristic value is calculated by formula (7).

Images