Drug stability test and evaluation method and device

Abstract

The invention relates to the technical field of drug data analysis, in particular to a drug stability test and evaluation method and device, which are used for collecting drug component data, analyzing the distribution and aggregation trend of drug components, evaluating the stability characteristics of the drug components under multiple conditions and generating a fractal dimension calculation result. According to the invention, by collecting and updating the drug component data in real time and evaluating the stability characteristics of the drug under variable environmental conditions, the environmental change can be reflected in time, and the timeliness and accuracy of the prediction result can be significantly improved. According to the method, the function of adaptively adjusting model parameters is achieved, and real-time environment data is combined, so that the prediction model can more accurately simulate behaviors of drugs under different environment conditions, and therefore, the testing flexibility is improved, and the processing capacity for complex data is enhanced. In addition, by analyzing the specific influence of environmental factors on the drug stability, the setting accuracy of drug storage and transportation conditions is further improved.

Description

Technical Field

[0001] The present invention relates to the technical field of drug data analysis, and in particular to a drug stability testing and evaluation method and device. Background Art

[0002] The field of drug data analysis technology optimizes drug R&D efficiency and improves drug safety assessment by processing and analyzing large amounts of data related to drug development, production, and application. The application of data analysis in drug R&D includes computational simulation of drug design, statistical analysis of clinical trial data, and data monitoring of drug side effects. By analyzing drug-related data, production decisions and market strategies can be made more accurately while ensuring drug quality and efficacy.

[0003] Among them, drug stability testing and evaluation methods use computational models and statistical techniques to evaluate the chemical and physical stability of drugs during storage and use. This evaluation is critical to determining the shelf life, storage conditions, and safe use of drugs. Using data analysis methods, a large amount of experimental data can be processed and interpreted to predict the performance of drugs under different environmental conditions.

[0004] Existing technologies are not flexible and timely enough when processing environmental data that may differ significantly from actual application conditions. In particular, in the face of rapidly changing environmental factors, it is difficult to effectively predict the impact of these changes on drug stability, resulting in drug storage guidelines and usage recommendations that may not be accurate enough. In addition, when large-scale data analysis is required to predict the stability of drugs under variable conditions, existing technologies have difficulty effectively processing and interpreting complex data, resulting in evaluation results that may not be accurate or timely enough. As a result, the setting of drug storage and transportation conditions is not accurate enough, affecting the effectiveness and safety of the drugs. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a drug stability testing and evaluation method and device.

[0006] In order to achieve the above object, the present invention adopts the following technical solution: a drug stability testing and evaluation method, comprising the following steps:

[0007] S1: Collect drug component data, analyze the distribution and aggregation trends of drug components, evaluate the stability characteristics of drug components under multiple conditions, and generate fractal dimension calculation results;

[0008] S2: Based on the fractal dimension calculation result, the current temperature, humidity and light intensity data are used as input to construct a model and set parameters to generate a drug stability prediction model;

[0009] S3: using the drug stability prediction model to simulate the behavior of the drug under predetermined environmental variables, identifying key factors that cause stability changes from the output of the model, and generating simulation test results;

[0010] S4: Based on the simulation test results, analyze the impact of key factors on drug stability, identify the effects of changes in temperature, humidity, and light on drug components, and generate key factor impact assessment results;

[0011] S5: Based on the evaluation results of the key factors, identify and analyze abnormal data points that show stability deviations, where the abnormal data points represent abnormal reactions of the drug components under specified environmental conditions, and generate abnormal detection results;

[0012] S6: According to the abnormal detection results, according to the current drug formulation or storage condition data, the response patterns of the drug components under various test environments are analyzed to obtain drug stability evaluation results.

[0013] The present invention has the following improvements: the fractal dimension calculation result includes the spatial diffusion coefficient, aggregation index and multi-condition environmental sensitivity score of the drug components; the drug stability prediction model includes the temperature sensitivity coefficient, humidity response parameter and light adjustment factor; the simulation test result includes the stability index under simulated temperature, humidity and light conditions, the drug decomposition rate and the adjusted warning threshold; the key factor impact assessment result includes the temperature change impact score, the humidity adjustment response value and the light intensity impact coefficient; the anomaly detection result includes the feature list of abnormally responsive drug components, the environmental stress anomaly index and the stability anomaly distribution map; the drug stability assessment result includes the environmental adaptability rating, the drug component stability index and the drug formula adjustment suggestion.

[0014] The present invention has improved the following specific steps of collecting drug component data, analyzing the distribution and aggregation trend of drug components, evaluating the stability characteristics of drug components under multiple conditions, and generating fractal dimension calculation results:

[0015] S101: Collect drug reaction data under various temperature, humidity and light conditions, format the data into a unified format, and generate an integrated data set;

[0016] S102: analyzing the distribution pattern of the drug components in the integrated data set, identifying the trends of aggregation and dispersion, evaluating the stability characteristics according to the trends, and generating stability trend analysis results;

[0017] S103: extracting data from the stability trend analysis results, calculating fractal dimension by box counting method, and generating fractal dimension calculation results according to the stability index of drug components under multiple conditions.

[0018] The present invention is improved in that the box counting method is according to the formula:

[0019]

[0020] Calculate the fractal dimension D f ;

[0021] Where ∈ represents the size of the box, N(∈) represents the minimum number of boxes required to cover the entire dataset at a given size ∈, and P temp is the temperature sensitivity coefficient, P hum is the humidity response coefficient, L intensity is the light intensity influence coefficient.

[0022] The present invention is improved in that, based on the fractal dimension calculation result, the current temperature, humidity and light intensity data are used as input to construct a model and set parameters, and the specific steps of generating a drug stability prediction model are as follows:

[0023] S201: Based on the fractal dimension calculation result, temperature, humidity and light intensity are selected as model input parameters, the basic structure of the drug stability prediction model is initialized, and an initial model structure is generated;

[0024] S202: adjusting and optimizing parameters of the initial model structure, simulating the influence of various storage conditions and environmental factors, performing multiple iterative tests, and generating a model iterative optimization result;

[0025] S203: Apply the iterative optimization result of the model to the model to reflect the change of drug stability under each environmental condition and generate a drug stability prediction model.

[0026] The present invention is improved in that the drug stability prediction model is used to simulate the behavior of the drug under predetermined environmental variables, and the key factors causing stability changes are identified from the output of the model. The specific steps of generating simulation test results are as follows:

[0027] S301: using the drug stability prediction model, inputting environmental variable data again, including real-time monitored temperature, humidity and light intensity, simulating the reaction of the drug under various conditions, and generating simulated input data;

[0028] S302: Analyze the simulation input data, identify the behavior changes of drug components under specified environmental variables, and the changes of stability indicators, and generate key behavior pattern analysis results;

[0029] S303: Based on the analysis results of the key behavior patterns, the responses of the drug components to the environmental variables are evaluated, key influencing factors are selected, including chemical activity changes caused by temperature changes, and simulation test results are generated.

[0030] The present invention is improved in that, according to the simulation test results, the influence of key factors on drug stability is analyzed, the effects of changes in temperature, humidity, and light on drug components are identified, and the specific steps of generating key factor influence evaluation results are as follows:

[0031] S401: Based on the simulation test results, perform an impact analysis of key environmental factors, measure the effects of temperature, humidity and light on the stability of drug components, and generate preliminary analysis results of the impact of environmental factors;

[0032] S402: extracting data from the preliminary analysis results of the environmental factors, calculating the quantitative effects of each key environmental factor including temperature, humidity and light on drug stability, and generating quantitative results of environmental factors;

[0033] S403: Based on the quantitative results of the environmental factors, the effects of various environmental parameters on the stability of drug components are compared, the core parameters that affect drug stability are selected, and the key factor impact assessment results are generated.

[0034] The present invention is improved in that, according to the key factors affecting the evaluation results, the abnormal data points showing stability deviation are identified and analyzed, and the abnormal data points characterize the abnormal reaction of the drug components under the specified environmental conditions. The specific steps of generating the abnormal detection results are as follows:

[0035] S501: Based on the key factor impact assessment result, screen the data points caused by extreme changes in environmental parameters, identify the data points as abnormal data points, and generate preliminary abnormal data point identification results;

[0036] S502: Analyze the environmental records in the preliminary identification results of the abnormal data points, compare with the conventional condition data, determine the relationship between the environmental changes and the abnormal reactions, and generate abnormal cause analysis results;

[0037] S503: Mark the abnormal data points identified in the abnormal cause analysis result, define a stability deviation feature for each marked point, and generate an abnormality detection result.

[0038] The present invention is improved in that, according to the abnormal detection results, according to the current drug formula or storage condition data, the response patterns of drug components under various test environments are analyzed to obtain the drug stability evaluation results in the following specific steps:

[0039] S601: Using the abnormal detection results, record and analyze the response pattern data under the drug formulation or storage conditions, determine the response pattern stability of the drug components, and generate a response pattern evaluation result;

[0040] S602: extracting data from the response mode evaluation results, comparing the stability performance of the drug component under each test environment, identifying the performance conditions in each case, and generating stability condition comparison results;

[0041] S603: Analyze the stability condition comparison results, summarize the performance of the drug components under the test environment, and generate drug stability evaluation results.

[0042] A drug stability testing and evaluation device, comprising:

[0043] The data collection and analysis module collects drug component information based on experimental data, analyzes the distribution and aggregation trends of drug components under various environmental conditions, evaluates the stability characteristics of drugs under various conditions, and generates fractal dimension calculation results;

[0044] The model building and simulation test module builds a drug stability prediction model based on the fractal dimension calculation results, uses environmental parameters to initialize the model and set parameters, simulates the behavior of the drug under set environmental conditions through the model, identifies the key factors that cause stability changes, and generates simulation test results;

[0045] The key factor analysis and evaluation module uses the simulation test results to analyze and quantify the impact of key environmental factors, including changes in temperature, humidity, and light, on the stability of drug ingredients, and generates key factor impact evaluation results;

[0046] The anomaly detection and processing module identifies and analyzes abnormal data points that show stability deviations based on the key factor impact assessment results, the data points identify abnormal reactions of drug ingredients under specified environmental conditions, and generates anomaly detection results;

[0047] The stability evaluation module evaluates the response patterns of the drug components in various test environments under the current drug formulation or storage conditions according to the abnormal detection results, and generates a drug stability evaluation result.

[0048] Compared with the prior art, the advantages and positive effects of the present invention are:

[0049] In the present invention, by collecting and updating drug ingredient data in real time and evaluating the stability characteristics of drugs under variable environmental conditions, environmental changes can be reflected immediately, significantly improving the timeliness and accuracy of prediction results. The function of adaptively adjusting model parameters, combined with real-time environmental data, enables the prediction model to more accurately simulate the behavior of drugs under different environmental conditions, thereby improving the flexibility of the test and enhancing the ability to handle complex data. In addition, by analyzing the specific impact of environmental factors on drug stability, the accuracy of setting drug storage and transportation conditions is further improved. The application of anomaly detection to identify stability deviations increases the ability to handle variables that affect drug effectiveness and safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 A flow chart of a drug stability testing and evaluation method is proposed for the present invention;

[0051] Figure 2 This is a detailed flow chart of step S1 of the present invention;

[0052] Figure 3 This is a detailed flow chart of step S2 of the present invention;

[0053] Figure 4 This is a detailed flow chart of step S3 of the present invention;

[0054] Figure 5 This is a detailed flow chart of step S4 of the present invention;

[0055] Figure 6 This is a detailed flow chart of step S5 of the present invention;

[0056] Figure 7 This is a detailed flow chart of step S6 of the present invention;

[0057] Figure 8 A module diagram of a drug stability testing and evaluation device is proposed for the present invention. DETAILED DESCRIPTION

[0058] In order 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 embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0059] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating positions or positional relationships, are based on the positions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, the meaning of "multiple" is two or more, unless otherwise clearly and specifically defined.

[0060] See also Figure 1 The present invention provides a technical solution: a method for testing and evaluating drug stability, comprising the following steps:

[0061] S1: Collect drug component data, analyze the distribution and aggregation trends of drug components, evaluate the stability characteristics of drug components under multiple conditions, and generate fractal dimension calculation results;

[0062] S2: Based on the fractal dimension calculation results, the current temperature, humidity and light intensity data are used as input to build the model and set parameters to generate a drug stability prediction model;

[0063] S3: Use the drug stability prediction model to simulate the behavior of the drug under predetermined environmental variables, identify the key factors that cause stability changes from the model output, and generate simulation test results;

[0064] S4: Based on the simulation test results, analyze the impact of key factors on drug stability, identify the effects of changes in temperature, humidity, and light on drug ingredients, and generate key factor impact assessment results;

[0065] S5: Based on the evaluation results of key factors, identify and analyze abnormal data points that show stability deviations. Abnormal data points represent abnormal reactions of drug ingredients under specified environmental conditions and generate abnormal detection results.

[0066] S6: Based on the abnormal detection results and the current drug formulation or storage condition data, the response patterns of the drug ingredients under various test environments are analyzed to obtain drug stability assessment results.

[0067] The fractal dimension calculation results include the spatial diffusion coefficient, aggregation index and multi-condition environmental sensitivity score of drug ingredients. The drug stability prediction model includes the temperature sensitivity coefficient, humidity response parameter and light regulation factor. The simulation test results include the stability index, drug decomposition rate and adjusted warning threshold under simulated temperature, humidity and light conditions. The key factor impact assessment results include the temperature change impact score, humidity regulation response value and light intensity impact coefficient. The anomaly detection results include the feature list of abnormally responding drug ingredients, the environmental stress anomaly index and the stability anomaly distribution map. The drug stability assessment results include the environmental adaptability rating, drug ingredient stability index and drug formulation adjustment suggestions.

[0068] See also Figure 2 , collect drug component data, analyze the distribution and aggregation trends of drug components, evaluate the stability characteristics of drug components under multiple conditions, and generate fractal dimension calculation results. The specific steps are as follows:

[0069] S101: Collect drug reaction data under various temperature, humidity and light conditions, format the data into a unified format, and generate an integrated data set;

[0070] To collect the reaction data of drugs under various temperature, humidity and light conditions, first set up the environmental simulation equipment in the laboratory and adjust it to different temperature levels (such as 20℃, 30℃, 40℃), humidity levels (such as 30%, 50%, 70%) and light intensities (such as 100Lux, 200Lux, 300Lux). During the experiment, the drug samples were exposed to these conditions, and the physical and chemical changes of the drugs were recorded regularly, including color, solubility, pH changes, and any observable decomposition or deterioration. After the data collection is completed, the obtained data is sorted and formatted to ensure that all data items follow a unified recording standard to facilitate subsequent data analysis and processing. For example, all temperature data are converted to degrees Celsius, humidity data are converted to percentages, and light intensity is converted to Lux units. At the same time, the timestamp and environmental conditions of each data point are recorded.

[0071] S102: Analyze the distribution pattern of drug components in the integrated data set, identify the trends of aggregation and dispersion, evaluate the stability characteristics based on the trends, and generate stability trend analysis results;

[0072] Among them, the distribution pattern of drug components in the integrated data set is analyzed according to the formula: Calculate the variance D of the drug component distribution. Where x i represents the concentration of the drug ingredient at the ith data point, μ represents the average concentration of these data points, and n represents the total number of data points.

[0073] Calculation process:

[0074] The distribution pattern analysis of drug components mainly focuses on evaluating the aggregation and dispersion trends of drug components under different environmental conditions. In order to quantify this distribution feature, the variance D is selected to evaluate drug stability. The variance calculation formula describes the average of the sum of the squares of the deviations of each data point from the mean, which can effectively express the degree of dispersion of the data.

[0075] Get each data point x through the actual monitoring data of drug ingredient concentration i , and then calculate the average value μ of these data. It is obtained during the experiment by the chemical analysis instrument and recorded in the database. Calculate the deviation between each data point and the average value, then square it, and finally divide the sum of all deviation squares by the total number of data points n to get the variance D.

[0076] For example, five data points were collected in the experiment, and the concentrations of the drug ingredient were 10 mg / L, 12 mg / L, 8 mg / L, 11 mg / L, and 9 mg / L, respectively.

[0077] Calculate the mean μ of these data as:

[0078] (10+12+8+11+9) / 5=10mg / L

[0079] Calculating the square of the deviation of each data point from the mean yields:

[0080] (0 2 +2 2 +(-2) 2 +1 2 +(-1) 2 )=10

[0081] Therefore, the variance D is 10 / 5 = 2 mg 2 / L 2 .

[0082] The results show that the dispersion of the drug components under these experimental conditions is relatively low, which may mean that the drug components are relatively stable under this set of experimental conditions. By further analyzing the relationship between the variance results and the stability of the drug, a more in-depth evaluation of the stability characteristics of the drug can be performed.

[0083] S103: extracting data from the stability trend analysis results, calculating the fractal dimension by a box counting method, and generating fractal dimension calculation results according to the stability index of the drug components under multiple conditions.

[0084] The box counting method follows the formula:

[0085]

[0086] Calculate the fractal dimension Df ;

[0087] Where ∈ represents the size of the box, N(∈) represents the minimum number of boxes required to cover the entire data set under a given size ∈, and the complexity of the data is evaluated by comparing the growth rate of the number of boxes N(∈) as ∈ shrinks. Log is a logarithmic function that converts the data growth rate to a comparable linear scale. P temp is the temperature sensitivity coefficient, P hum is the humidity response coefficient, L intensity is the light intensity influence coefficient, obtained through data analysis, based on the sensitive response of drug components to environmental variables in the experiment;

[0088] The calculation process is as follows:

[0089] Assume that in a certain experimental setting, it is observed that when ∈ = 0.1, the number of boxes N(∈) without considering environmental factors is 1500. By analyzing the reaction data of the drug under different environmental conditions, the temperature sensitivity coefficient P is determined. temp =1.05, indicating that when the temperature rises by 5%, the drug stability decreases by 5%. Humidity response coefficient P hum =0.95, indicating that for every 5% increase in humidity, the drug stability increases by 5%. Light intensity influence coefficient L intensity =1.1, indicating that for every 10% increase in light intensity, the drug stability decreases by 10%.

[0090] Substitute these values ​​into the improved formula for calculation:

[0091]

[0092]

[0093] D f =log(1500·1.05·0.95·1.1)

[0094] D f =log(1645.88)≈3.22

[0095] The results show that the fractal dimension after the introduction of environmental parameters is 3.22, which provides a more accurate measure of data complexity than the previous value without the introduction of parameters, reflecting the complexity of the spatial distribution of drug components under the influence of different environmental factors. The increase in this value indicates that drug components show more complex distribution characteristics after considering the influence of temperature, humidity and light.

[0096] See also Figure 3 Based on the fractal dimension calculation results, the current temperature, humidity and light intensity data are used as input to build the model and set parameters. The specific steps to generate the drug stability prediction model are as follows:

[0097] S201: Based on the fractal dimension calculation results, temperature, humidity and light intensity are selected as model input parameters, the basic structure of the drug stability prediction model is initialized, and an initial model structure is generated;

[0098] Based on the results of fractal dimension calculation, temperature, humidity and light intensity are selected as model input parameters. This step first requires evaluating the fractal dimension calculation results obtained from the previous analysis to determine how the three parameters of temperature, humidity and light intensity affect the stability of drug ingredients. By deeply analyzing the reaction data of drug ingredients under different temperature, humidity and light conditions, it is determined how changes in these environmental factors affect the physical and chemical properties of drugs. For example, high temperature may accelerate the decomposition of drug ingredients, while high humidity may affect the solubility or hygroscopicity of drugs. Next, the infrastructure of the drug stability prediction model is initialized, which involves selecting the appropriate model type to build a prediction framework. Depending on the characteristics of the drug ingredients and the accuracy of the required prediction, a linear model is selected to handle direct relationships, or a machine learning model such as a neural network is used to handle nonlinear relationships and a large number of input variables. At this stage, various parameters and hierarchical structures of the model need to be set, such as how many hidden layers are set in the neural network model and the number of nodes in each layer. These settings are based on the results of previous experiments and data analysis. In addition, in order to enable the model to effectively process the input environmental parameters, parameter preprocessing can be performed, which includes standardizing or normalizing the temperature, humidity, and light intensity data to eliminate the impact of different magnitudes and dimensions, ensuring the consistency of data input and the efficiency of model training. Finally, generating the initial model structure includes a complete data processing flow, model architecture, and initial training configuration.

[0099] S202: adjusting and optimizing parameters of the initial model structure, simulating the influence of various storage conditions and environmental factors, performing multiple iterative tests, and generating model iterative optimization results;

[0100] First, a simulation test framework is needed that includes multiple combinations of environmental variables that simulate different storage conditions, such as different temperatures, humidity, and light intensities, as well as their interactions. In specific implementation, a baseline data set is established through the collected drug stability data, which includes the stability performance of the drug under standard conditions. Next, for each parameter in the model, such as weights, biases, and network layer configurations, the performance of the model under different environmental combinations is evaluated by using a cross-validation method. These parameters are adjusted through actual simulated data to reduce prediction errors. To this end, automated parameter adjustments, such as grid search or random search, are used to explore the effects of different parameter combinations. In each round of iteration, the model is trained with a new combination of parameters and its performance is tested on the validation set. During the iteration process, the performance of each set of parameters and the response of the model to environmental changes are recorded. In addition, the test simulating multiple storage conditions is not limited to existing data, but can also be expanded by generating synthetic data to more comprehensively evaluate the stability and adaptability of the model. After each iteration, the deviation of the model output from the expected results is analyzed, and the model structure and parameters are refined based on these deviations. Finally, the model is finally optimized based on the iterative test results. This involves solidifying those parameter settings that perform best and removing those that perform poorly.

[0101] S203: Apply the model iterative optimization results to the model to reflect the changes in drug stability under each environmental condition and generate a drug stability prediction model.

[0102] First, based on the results of previous iterative tests, the best performing model parameter configuration is selected, including network structure, learning rate, number of layers, and activation function. These parameters have been verified to provide the most accurate prediction results under a variety of simulated environmental conditions. Next, these parameters are fixed and loaded into the model's infrastructure to form a complete drug stability prediction model. Subsequently, a complete training dataset is used, including drug stability data collected from laboratories and actual applications. During the training process, the model will learn how to predict the stability of drugs based on input parameters such as temperature, humidity, and light intensity. In order to ensure the generalization ability of the model, an independent validation dataset is used to monitor the performance during the training process and make necessary adjustments. After the training is completed, a series of validation tests are conducted, which use data that was not previously involved in the training process to evaluate the model's prediction accuracy under new environmental conditions. These tests can confirm whether the model can accurately reflect the impact of various environmental factors on drug stability.

[0103] See also Figure 4 , using the drug stability prediction model, simulate the behavior of the drug under predetermined environmental variables, identify the key factors that cause stability changes from the output of the model, and generate the simulation test results in the following specific steps:

[0104] S301: using the drug stability prediction model, inputting environmental variable data again, including real-time monitored temperature, humidity and light intensity, simulating the reaction of the drug under various conditions, and generating simulated input data;

[0105] Using the optimized drug stability prediction model, the real-time monitored environmental variable data are re-entered into the model for simulation testing. These environmental variables include temperature, humidity and light intensity, which are collected in real time from the laboratory or actual storage environment. The collected data need to be processed initially, such as removing noise, filling missing values ​​and performing necessary data formatting to ensure that they are suitable for input into the prediction model. The processed environmental variable data are input into the drug stability prediction model. The model uses these real-time data as input parameters to simulate the reaction of the drug under current and future environmental conditions. This includes predicting the decomposition rate of the drug, possible chemical changes, and its stability indicators. In this process, the model uses the random forest algorithm to predict the behavior of the drug based on the input environmental data. It can also simulate the impact of extreme high temperature and high humidity environments on drug stability, or evaluate the response of drug ingredients under unstable light conditions. Understand the stability performance of the drug under various possible storage conditions, and then guide the optimization of packaging and storage strategies.

[0106] S302: Analyze simulation input data, identify behavioral changes of drug components under specified environmental variables, and changes in stability indicators, and generate key behavioral pattern analysis results;

[0107] Centrally process and evaluate the simulation results obtained from the random forest model, including the behavior data of drug ingredients under different environmental variables. The simulation results are analyzed by data analysis software to identify the changes in the behavior of drug ingredients and changes in stability indicators under various environmental conditions. The analysis process includes comparing the changes in drug stability under different environmental conditions (such as different temperatures, humidity and light intensity), and identifying the environmental variables that are most relevant to stability changes. In addition, statistical methods and visualization tools such as scatter plots and heat maps are used to reveal key patterns and trends in the behavior of drug ingredients. Summarize the key behavioral patterns of drug ingredients under specific conditions.

[0108] S303: Based on the results of the key behavior pattern analysis, evaluate the response of the drug ingredients to the environmental variables, select key influencing factors, including changes in chemical activity caused by temperature changes, and generate simulation test results.

[0109] Based on the results of the key behavior pattern analysis, the response of drug ingredients to environmental variables is evaluated. The key behavior patterns obtained from data analysis are used to determine the main environmental factors that affect drug stability, such as the effect of temperature changes on drug chemical activity. For example, by analyzing chemical activity changes such as accelerated drug decomposition caused by temperature rise, key influencing factors are selected and these factors are quantitatively evaluated more accurately. Finally, based on these evaluation results, simulation test results are generated to record how each environmental factor affects the stability of drug ingredients.

[0110] See also Figure 5 , based on the simulation test results, analyze the impact of key factors on drug stability, identify the effects of changes in temperature, humidity, and light on drug ingredients, and generate key factor impact assessment results. The specific steps are as follows:

[0111] S401: Based on the simulation test results, conduct an impact analysis of key environmental factors, measure the effects of temperature, humidity and light on the stability of drug ingredients, and generate preliminary analysis results of the impact of environmental factors;

[0112] Among them, the simulation test results are analyzed according to the formula ΔS = c T ΔT+c H ·ΔH+c L ·ΔL, calculate the change in drug stability. In the formula, ΔS represents the change in drug stability, ΔT represents the temperature change, ΔH represents the humidity change, ΔL represents the light intensity change, c T 、c H and c L are the coefficients of influence of temperature, humidity and light on stability.

[0113] When conducting the impact analysis of key environmental factors, the data collected from the simulation test results are used to evaluate the specific effects of different environmental variables (temperature, humidity, and light) on drug stability. A linear relationship model is established to quantify the influence of each environmental factor.

[0114] First, the coefficient c is determined based on experimental data T 、c H and c L The coefficients are obtained through regression analysis based on the current data set, which contains observations of drug stability under different environmental conditions. For example, if the drug stability decreases by 0.5% for every 1°C increase in temperature, then c T =-0.5.

[0115] In a set of simulation tests, the temperature increased from 20℃ to 25℃, the humidity increased from 50% to 60%, and the light increased from 200Lux to 250Lux. According to the known coefficient c T =-0.5, cH = -0.3 and c L =-0.2, calculate the change in drug stability:

[0116] ΔT=25-20=5

[0117] ΔH=60-50=10

[0118] ΔL=250-200=50

[0119] ΔS=(-0.5·5)+(-0.3·10)+(-0.2·50)

[0120] ΔS=-2.5-3-10

[0121] ΔS=-15.5

[0122] The results show that considering the change of given environmental variables, the drug stability is expected to decrease by 15.5%. It helps to understand the specific impact of specific environmental changes on drug stability and make corresponding storage and management adjustments in practical applications.

[0123] S402: extracting data from the preliminary analysis results of environmental factors, calculating the quantitative effect of each key environmental factor including temperature, humidity and light on drug stability, and generating quantitative results of environmental factors;

[0124] Among them, data is extracted from the preliminary analysis results of environmental factors, according to the formula I = ∑ i′ n′ =1a i′ b i′ , calculate the quantitative impact of each key environmental factor. In the formula, I represents the total environmental factor impact value, a i′ represents the change in the i′th environmental factor (such as temperature, humidity and light), b i′ represents the weight of the corresponding environmental factor, n′ is the total number of environmental factors, and represents the number of different types of environmental variables considered in the analysis.

[0125] By quantifying the impact of key environmental factors on drug stability, n' refers to the number of environmental factors considered in the model. For example, if the analysis includes temperature, humidity, and light intensity, then n' = 3. In this way, each environmental factor has a corresponding change a i′ and weight b i′ , these variables make it possible to calculate how changes in environmental conditions affect the stability of the drug.

[0126] The setting parameters determine the weights of the environmental factors temperature, humidity and light, respectively. T′ =0.4, b H′ =0.35, bL′ =0.25. If the temperature change is 5°C, the humidity change is 10%, and the light change is 50Lux, then:

[0127] a T′ =5,a H′ =10,a L′ =50

[0128] I=(0.4·5)+(0.35·10)+(0.25·50)

[0129] I=2+3.5+12.5

[0130] I=18

[0131] The calculation result I=18 shows that after considering the weights of each factor, the combined impact of environmental changes on drug stability is 18 units. This result helps to more fully understand the combined effects of different environmental factors, so as to better optimize the storage conditions and management strategies of drugs. If it is known from previous stability tests that under safe and stable conditions, the combined impact of environmental factors usually does not exceed 10. In new tests or actual applications, if the calculated impact value of 18 exceeds the baseline value, it indicates that the environmental conditions may have a significant negative impact on drug stability, and additional control measures or formulation adjustments are required. In this way, the baseline value serves as an important reference point to help determine whether the current environmental settings are appropriate or whether improvements are needed to ensure the stability and efficacy of the drug.

[0132] S403: Based on the quantitative results of the impact of environmental factors, compare the impact of various environmental parameters on the stability of drug ingredients, select the core parameters that affect drug stability, and generate key factor impact assessment results.

[0133] Among them, based on the quantitative results of environmental factors, the effects of various environmental parameters on the stability of drug components were compared, and the formula R = max (a i ′ b i ′ ) calculation, and the core parameters affecting drug stability are selected. In the formula, R represents the maximum single environmental factor affecting drug stability, a i ′ and b i ′ are the change amount and weight of environmental factors respectively, where i′ represents the index of each environmental factor.

[0134] In order to identify which environmental factor has the most significant impact on drug stability in the current model and data analysis, the combined influence of each factor can be quantified by calculating the product of the change in each environmental factor and its weight, and the single environmental factor with the greatest impact on drug stability can be determined by taking the maximum value R of these product values.

[0135] In the previous analysis, the change amount and weight of the environmental factors temperature, humidity and light are known to be a T ′ =5,b T ′ =0.4, a H ′ =10,b H ′ =0.35, a L ′ =50,b L ′ =0.25.

[0136] Calculate the impact value of each factor:

[0137] Temperature effect = 5 0.4 = 2

[0138] Humidity effect = 10 0.35 = 3.5

[0139] Light impact = 50 0.25 = 12.5

[0140] R=max(2,3.5,12.5)=12.5

[0141] The calculated result R = 12.5 indicates that among the environmental variables considered, light changes have the most significant impact on drug stability. The results guide the R&D team to pay special attention to the control of light conditions during drug storage and management to maximize the maintenance of drug stability.

[0142] See also Figure 6 , according to the key factors affecting the evaluation results, identify and analyze abnormal data points that show stability deviations. Abnormal data points represent abnormal reactions of drug ingredients under specified environmental conditions. The specific steps for generating abnormal detection results are as follows:

[0143] S501: Based on the key factor impact assessment results, screen the data points caused by extreme changes in environmental parameters, identify the data points as abnormal data points, and generate preliminary identification results of abnormal data points;

[0144] Among them, the data points caused by extreme changes in environmental parameters are screened according to the formula: E d =∑ k＝ m1|v k -μ k|, calculate the total deviation of environmental parameters, compare it with the threshold θ to identify abnormal data points, and generate preliminary identification results of abnormal data points. d represents the total deviation of environmental parameters, v k represents the observed value of the kth environmental parameter, μ k represents the average value of the kth environmental parameter, θ represents the set abnormal threshold, and m represents the total number of environmental parameters.

[0145] To quantify the deviation between the observed value of each environmental parameter and its mean value, and to sum it up to E d The overall deviation is evaluated by calculating the value of the parameter. This detects data points that may become abnormal due to extreme changes in environmental parameters.

[0146] Set three environmental parameters: temperature (v temp ), humidity (v hum ), light (v light ), whose average values ​​are set as μ temp =25℃,μ hum =50%,μ light = 200 Lux. Set the abnormal threshold θ = 10. For example, in a certain test, the observation is v temp =40℃,v hum =60%, v light

[0147] =250Lux

[0148] Calculate the sum of deviations:

[0149] E d =|40-25|+|60-50|+|250-200|

[0150] E d =15+10+50

[0151] E d =75

[0152] The calculated deviation sum E d =75 is compared with the threshold value θ=10. d >θ, indicating that the data points in the test were identified as abnormal due to extreme changes in environmental parameters. This result shows that significant changes in temperature and light, in particular, need to be considered as key environmental factors that may have an impact on drug stability during monitoring.

[0153] S502: Analyze the environmental records in the preliminary identification results of abnormal data points, compare with conventional condition data, determine the relationship between environmental changes and abnormal reactions, and generate abnormal cause analysis results;

[0154] Among them, the environmental records in the preliminary identification results of abnormal data points are analyzed according to the formula Calculate the standard deviation of environmental parameter changes, and then determine the relationship between environmental changes and abnormal reactions, and generate abnormal cause analysis results. In the formula, ΔE represents the standard deviation of environmental parameter changes, v k represents the observed value of the kth environmental parameter, μ k represents the average value of the kth environmental parameter, and m represents the total number of environmental parameters.

[0155] Using standard deviation to quantify the degree of deviation between each environmental parameter and its mean value can help reveal the potential connection between environmental changes and abnormal drug reactions. Standard deviation is a statistical tool to measure the degree of dispersion of data points, and the calculated results can help determine the statistical significance of changes in environmental parameters.

[0156] Continue to use the above environmental parameters: temperature (v temp =40℃), humidity (v hum =60%), light (v light =250Lux), and their respective average values ​​μ temp =25℃,μ hum =50%,μ light

[0157] =200Lux. Calculate the standard deviation:

[0158]

[0159] ΔE≈53.15

[0160] The result ΔE≈53.15 shows that the environmental parameters have a large deviation from the average value, indicating obvious environmental fluctuations, which may be the direct cause of abnormal drug stability. Through further analysis, it is possible to specifically indicate which environmental parameters have changed most dramatically, and take measures accordingly to prevent similar drug stability problems in the future.

[0161] S503: Mark the abnormal data points identified in the abnormal cause analysis result, define stability deviation characteristics for each marked point, and generate an abnormality detection result.

[0162] Among them, the abnormal data points identified in the abnormal cause analysis results are marked, and the formula is used Define stability deviation features for each marked point and generate anomaly detection results. Where F represents the total stability deviation feature score, y j Represents the stability deviation of the jth marker point, z j Represents the corresponding weight factor, and p represents the total number of marked points.

[0163] Quantify and describe the stability deviation of each data point identified as an anomaly. By multiplying the stability deviation of each flagged point with a weighting factor, a composite score is generated, which helps to further evaluate and prioritize these anomalies, providing a basis for possible corrective actions.

[0164] There are three marked abnormal data points, with stability deviations of y1=3.5, y2=2.0, y3=5.0, and corresponding weight factors z1=0.5, z2=1.0, z3=0.8. Calculate the total stability deviation feature score:

[0165] F=(3.5·0.5)+(2.0·1.0)+(5.0·0.8)

[0166] F=1.75+2.0+4.0

[0167] F=7.75

[0168] The result F = 7.75 shows the cumulative stability deviation feature score, indicating that among these marked abnormal data points, the third point has the greatest impact on the overall stability and may need priority attention. This helps understand which specific conditions require additional attention and may guide further stability improvements or adjust storage and handling conditions to reduce future abnormalities.

[0169] See also Figure 7 ,According to the abnormal detection results, according to the current drug formulation or storage condition data, the response patterns of drug ingredients under various test environments are analyzed, and the specific steps to obtain the drug stability evaluation results are as follows:

[0170] S601: Using the abnormal detection results, record and analyze the response pattern data under the drug formulation or storage conditions, determine the response pattern stability of the drug components, and generate a response pattern evaluation result;

[0171] Among them, the reaction pattern data of drug formulation or storage conditions are recorded and analyzed according to the formula Calculate the variance of the drug component response pattern and generate the response pattern evaluation result. r represents the variance of the response pattern of the drug component, r h represents the drug response data for the hth test point, represents the average of drug response data of all test points, and g represents the total number of test points.

[0172] Evaluate the stability of the response pattern of a drug ingredient under different formulations or storage conditions. By calculating the variance of drug response data, the consistency and stability of the drug ingredient response can be quantified. Variance Vr It reveals the degree of dispersion among the data points, with lower variance indicating a more stable response to the drug ingredient, while higher variance indicates greater variability in the response.

[0173] There are 5 test points of drug response data r = [15, 18, 15, 14, 17], calculate the average drug response:

[0174]

[0175] Calculate the variance:

[0176]

[0177] Results r =2.16 represents the variance of the drug component response at different test points. If historical data shows that the variance of similar drugs under similar test conditions is usually in the range of 1.0 to 3.0, then the variance value of 2.16 can be considered to be within the normal range, indicating that the drug component response is within the acceptable stability range. This indicates that the drug performance is relatively consistent, without significant abnormal fluctuations.

[0178] S602: extracting data from the response pattern evaluation results, comparing the stability performance of the drug component under each test environment, identifying the performance conditions under each condition, and generating stability condition comparison results;

[0179] Among them, data is extracted from the response mode evaluation results according to the formula D′ c =max(s h )-min(s h ), calculate the range difference of drug component stability performance under each test environment, and generate stability condition comparison results. c Represents the maximum difference in the stability performance of drug ingredients, s h Represents the drug stability score under the hth test environment.

[0180] Quantify and compare the stability performance of a drug ingredient under different testing conditions, and evaluate the extent of variation in performance conditions by calculating the maximum difference in the drug stability score under each environment. This calculation helps reveal the environment in which the drug is most or least stable under a given storage or formulation condition.

[0181] In five different test environments, the drug stability scores are s = [82, 88, 90, 85, 80]. First, find the maximum and minimum scores:

[0182] max(s h )=90

[0183] min(s h )=80

[0184] Compute the difference in stability conditions:

[0185] D′ c =max(s h )-min(s h )=90-80=10

[0186] The calculation result D′ c =10 means that the maximum difference in drug stability performance under the considered test environment is 10 points. The difference shows that there is a certain fluctuation in the stability performance of the drug under different environments. Further analysis of these differences will help optimize the formulation or storage conditions of the drug to improve its stability performance under various environments.

[0187] S603: Analyze the stability condition comparison results, summarize the performance of the drug components under the test environment, and generate drug stability evaluation results.

[0188] Among them, the stability condition comparison results are analyzed according to the formula Calculate the average stability score of the drug components under the test environment to generate the drug stability evaluation results. f Represents the average stability score of the drug component under the test environment, s h′ represents the drug stability score under the h′th test environment, and p represents the total number of test environments.

[0189] By calculating the average of the drug stability scores under all test environments, a comprehensive perspective is provided to evaluate the overall stability performance of the drug under different conditions. The calculation of this average score helps to fully understand the behavior of drug ingredients in diverse test environments and provide a basis for formulating effective quality control and improvement measures.

[0190] In five different test environments, the drug stability scores are s = [82, 88, 90, 85, 80], where each score corresponds to the environment h′ = 1 to h′ = 5.

[0191] Compute the average of these scores:

[0192]

[0193] S f =85

[0194] The calculation result S f=85 shows that the average stability score of the drug ingredients under all test conditions is 85. The average value reveals that the drug exhibits good overall stability in the current set of test conditions, indicating that the drug has good performance consistency and supports use or storage under these conditions. It helps to further guide the optimization of drug formulation or the adjustment of storage conditions to ensure drug quality and efficacy.

[0195] See also Figure 8 , a drug stability testing and evaluation device, the device comprising:

[0196] The data collection and analysis module collects drug component information based on experimental data, analyzes the distribution and aggregation trends of drug components under various environmental conditions, evaluates the stability characteristics of drugs under various conditions, and generates fractal dimension calculation results;

[0197] The model building and simulation test module builds a drug stability prediction model based on the fractal dimension calculation results, uses environmental parameters to initialize the model and set parameters, simulates the behavior of the drug under set environmental conditions through the model, identifies the key factors that cause stability changes, and generates simulation test results;

[0198] The key factor analysis and evaluation module uses simulation test results to analyze and quantify the impact of key environmental factors, including changes in temperature, humidity, and light, on the stability of drug ingredients, and generates key factor impact evaluation results;

[0199] The anomaly detection and processing module identifies and analyzes abnormal data points that show stability deviations based on the key factors affecting the evaluation results. The data points identify abnormal reactions of drug ingredients under specified environmental conditions and generate anomaly detection results.

[0200] The stability assessment module evaluates the response patterns of drug ingredients in multiple test environments under the current drug formulation or storage conditions based on the abnormal detection results, and generates drug stability assessment results.

[0201] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A drug stability testing and evaluation method, characterized in that: The following steps are involved: Collect drug ingredient data, analyze the distribution and aggregation trends of drug ingredients, evaluate the stability characteristics of drug ingredients under multiple conditions, and generate fractal dimension calculation results; Based on the fractal dimension calculation results, current temperature, humidity and light intensity data are used as input to construct a model and set parameters to generate a drug stability prediction model; Using the drug stability prediction model, simulating the behavior of the drug under predetermined environmental variables, identifying key factors that cause stability changes from the output of the model, and generating simulation test results; Based on the simulation test results, analyze the impact of key factors on drug stability, identify the effects of changes in temperature, humidity, and light on drug ingredients, and generate key factor impact assessment results; Based on the evaluation results of the key factors, identifying and analyzing abnormal data points showing stability deviations, the abnormal data points representing abnormal reactions of drug ingredients under specified environmental conditions, and generating abnormal detection results; According to the abnormal detection results, based on the current drug formulation or storage condition data, the response patterns of the drug components under various test environments are analyzed to obtain drug stability evaluation results.

2. The drug stability testing and evaluation method according to claim 1, characterized in that: The fractal dimension calculation results include the spatial diffusion coefficient, aggregation index and multi-condition environmental sensitivity score of the drug ingredients, the drug stability prediction model includes the temperature sensitivity coefficient, humidity response parameter and light adjustment factor, the simulation test results include the stability index under simulated temperature, humidity and light conditions, the drug decomposition rate and the adjusted warning threshold, the key factor impact assessment results include the temperature change impact score, the humidity adjustment response value and the light intensity impact coefficient, the anomaly detection results include the feature list of abnormally responsive drug ingredients, the environmental stress anomaly index and the stability anomaly distribution map, and the drug stability assessment results include the environmental adaptability rating, the drug ingredient stability index and the drug formula adjustment suggestion.

3. The drug stability testing and evaluation method according to claim 1, characterized in that: The specific steps for collecting drug component data, analyzing the distribution and aggregation trends of drug components, evaluating the stability characteristics of drug components under multiple conditions, and generating fractal dimension calculation results are as follows: Collect drug reaction data under various temperature, humidity and light conditions, format the data into a unified format, and generate an integrated data set; Analyzing the distribution patterns of drug components in the integrated data set, identifying trends of aggregation and dispersion, evaluating stability characteristics based on the trends, and generating stability trend analysis results; Data is extracted from the stability trend analysis results, and the fractal dimension is calculated by a box counting method. The fractal dimension calculation results are generated according to the stability index of the drug component under multiple conditions.

4. The drug stability testing and evaluation method according to claim 3, characterized in that: The box counting method follows the formula: Calculate the fractal dimension D f ; Where ∈ represents the size of the box, N(∈) represents the minimum number of boxes required to cover the entire dataset at a given size ∈, and P temp is the temperature sensitivity coefficient, P hum is the humidity response coefficient, L intensity is the light intensity influence coefficient.

5. The drug stability testing and evaluation method according to claim 1, characterized in that: Based on the fractal dimension calculation results, the current temperature, humidity and light intensity data are used as input to build the model and set parameters. The specific steps for generating the drug stability prediction model are as follows: Based on the fractal dimension calculation results, temperature, humidity and light intensity are selected as model input parameters, the basic structure of the drug stability prediction model is initialized, and an initial model structure is generated; Adjusting and optimizing the parameters of the initial model structure, performing multiple iteration tests by simulating the influence of various storage conditions and environmental factors, and generating model iteration optimization results; The iterative optimization results of the model are applied to the model to reflect the changes in drug stability under each environmental condition and generate a drug stability prediction model.

6. The drug stability testing and evaluation method according to claim 1, characterized in that: The drug stability prediction model is used to simulate the behavior of the drug under predetermined environmental variables, and the key factors that cause stability changes are identified from the output of the model. The specific steps for generating simulation test results are as follows: Using the drug stability prediction model, inputting environmental variable data again, including real-time monitored temperature, humidity and light intensity, simulating the reaction of the drug under various conditions, and generating simulated input data; Analyzing the simulation input data, identifying behavioral changes of drug components and changes in stability indicators under specified environmental variables, and generating key behavioral pattern analysis results; Based on the analysis results of the key behavior patterns, the responses of drug ingredients to environmental variables are evaluated, key influencing factors are selected, including changes in chemical activity caused by temperature changes, and simulation test results are generated.

7. The drug stability testing and evaluation method according to claim 1, characterized in that: According to the simulation test results, the specific steps for analyzing the impact of key factors on drug stability, identifying the effects of changes in temperature, humidity, and light on drug components, and generating key factor impact assessment results are as follows: Based on the simulation test results, perform an impact analysis of key environmental factors, measure the effects of temperature, humidity and light on the stability of drug ingredients, and generate preliminary analysis results of the impact of environmental factors; Extracting data from the preliminary analysis results of the environmental factors, calculating the quantitative effects of each key environmental factor including temperature, humidity and light on drug stability, and generating quantitative results of environmental factors; Based on the quantitative results of the environmental factors, the effects of various environmental parameters on the stability of drug ingredients are compared, the core parameters that affect drug stability are selected, and the key factor impact assessment results are generated.

8. The drug stability testing and evaluation method according to claim 1, characterized in that: According to the evaluation results of the key factors, the abnormal data points showing stability deviation are identified and analyzed. The abnormal data points represent the abnormal reaction of the drug ingredients under the specified environmental conditions. The specific steps for generating the abnormal detection results are as follows: Based on the impact assessment results of the key factors, data points caused by extreme changes in environmental parameters are screened, the data points are identified as abnormal data points, and preliminary identification results of abnormal data points are generated; Analyze the environmental records in the preliminary identification results of the abnormal data points, compare with the conventional condition data, determine the relationship between the environmental changes and the abnormal reactions, and generate abnormal cause analysis results; The abnormal data points identified in the abnormal cause analysis result are marked, and a stability deviation feature is defined for each marked point to generate an abnormality detection result.

9. The drug stability testing and evaluation method according to claim 1, characterized in that: According to the abnormal detection results, the specific steps of analyzing the response patterns of drug components under various test environments and obtaining drug stability evaluation results according to the current drug formulation or storage condition data are as follows: Using the abnormal detection results, recording and analyzing the response pattern data under the drug formulation or storage conditions, determining the response pattern stability of the drug components, and generating response pattern evaluation results; extracting data from the response pattern evaluation results, comparing the stability performance of the drug substance under each test environment, identifying the performance conditions under each condition, and generating stability condition comparison results; The stability condition comparison results are analyzed, the performance of the drug components under the test environment is summarized, and the drug stability evaluation results are generated.

10. A drug stability testing and evaluation device, characterized in that: According to any one of claims 1 to 9, the drug stability testing and evaluation method is implemented, and the device comprises: The data collection and analysis module collects drug component information based on experimental data, analyzes the distribution and aggregation trends of drug components under various environmental conditions, evaluates the stability characteristics of drugs under various conditions, and generates fractal dimension calculation results; The model building and simulation test module builds a drug stability prediction model based on the fractal dimension calculation results, uses environmental parameters to initialize the model and set parameters, simulates the behavior of the drug under set environmental conditions through the model, identifies the key factors that cause stability changes, and generates simulation test results; The key factor analysis and evaluation module uses the simulation test results to analyze and quantify the impact of key environmental factors, including changes in temperature, humidity, and light, on the stability of drug ingredients, and generates key factor impact evaluation results; The anomaly detection and processing module identifies and analyzes abnormal data points that show stability deviations based on the key factor impact assessment results, the data points identify abnormal reactions of drug ingredients under specified environmental conditions, and generates anomaly detection results; The stability evaluation module evaluates the response patterns of the drug components in various test environments under the current drug formulation or storage conditions according to the abnormal detection results, and generates a drug stability evaluation result.

Images