Drug granulation process optimization method and system based on artificial intelligence

Abstract

The invention relates to the technical field of medicine wet granulation, in particular to a medicine granulation process optimization method and system based on artificial intelligence, and the method comprises the steps: measuring material attribute data in a material stage, and obtaining material attribute parameter information; measuring airflow data and spraying data in a spraying liquid adding stage to obtain first process parameter information; measuring airflow data and spray data in a particle growth stage to obtain second process parameter information; analyzing the first process parameter information and the second process parameter information to obtain a process parameter analysis result; the dissolution characteristic of the prepared particles is measured, and dissolution information of the prepared particles is obtained; inputting the material attribute parameter information, the process parameter analysis result and the prepared particle dissolution information into a process optimization model to obtain process adjustment parameters; and adaptively adjusting the parameters of any one of the material stage, the spraying and liquid adding stage and the particle growth stage according to the process adjusting parameters. According to the method and the system disclosed by the invention, the consistency and the stability of the prepared particles are ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of drug wet granulation, and in particular to an artificial intelligence-based drug granulation process optimization method and system. Background Art

[0002] Fluidized bed granulation (also known as boiling granulation) is a granulation method that uses airflow from bottom to top to suspend drug powder in a fluidized state, and sprays liquid binder to agglomerate the powder into granules. During operation, the material powder is loaded into the material container, and airflow of suitable temperature is blown from the bottom of the bed layer to mix the material evenly in a fluidized state, and then liquid binder is sprayed to agglomerate the powder into granules (spraying and liquid addition stage). After repeated spraying and drying, spraying is stopped until the particles grow to the required size (particle growth stage), and the formed particles continue to be dried by hot air in the bed layer until the moisture content meets the requirements.

[0003] The dissolution characteristics of granulated drugs are important indicators for evaluating the release and absorption efficiency of drugs in vivo. The airflow and spray during fluidization and growth are related to the dissolution characteristics of the produced granules, and the material properties at the material stage (including material density, material viscosity, material moisture content, and material / excipient ratio, etc.) are related to the dissolution characteristics of the produced granules. Granules produced from the same batch of powder raw materials and materials often have similar dissolution characteristics. When faced with different batches of powder raw materials and materials of the same type of drug, the dissolution characteristics of the produced granules may be significantly different. Currently, most pharmaceutical companies conduct dissolution tests during the process development stage, but in the actual continuous production process, there is a lack of real-time monitoring of the dissolution characteristics of particles under dynamic production conditions, which may result in products that do not meet drug standards. Summary of the invention

[0004] To this end, the technical problem to be solved by the present invention is to overcome the quality problems of drug granulation caused by deviations in dissolution characteristics in the prior art, and to provide a drug granulation process optimization method and system based on artificial intelligence to ensure that the granules produced from different batches of powder raw materials and materials of the same model have consistency and stability.

[0005] In the first aspect, in order to solve the above technical problems, the present invention provides a method for optimizing drug granulation process based on artificial intelligence, wherein drug granulation includes a material stage, a spraying liquid addition stage and a particle growth stage, and the method includes:

[0006] Determine the material attribute data at the material stage to obtain material attribute parameter information;

[0007] Determine the air flow data and the spray data in the spray liquid addition stage to obtain first process parameter information;

[0008] measuring the air flow data and the spray data in the particle growth stage to obtain second process parameter information;

[0009] Analyze the first process parameter information and the second process parameter information to obtain a process parameter analysis result;

[0010] Determine the dissolution characteristics of the prepared granules to obtain dissolution information of the prepared granules;

[0011] Inputting the material attribute parameter information, the process parameter analysis results and the dissolution information of the produced particles into a process optimization model to obtain process adjustment parameters;

[0012] According to the process adjustment parameters, the parameters of any of the three stages, namely, the material stage, the spraying and liquid addition stage, and the particle growth stage, are adaptively adjusted.

[0013] In one embodiment of the present invention, the air flow data and the spray data in the spray liquid addition stage are measured to obtain the first process parameter information, including:

[0014] Divide the area according to the structure of the airflow screen plate, set a test point in each area, collect the airflow velocity and airflow temperature of the test point, and obtain the first-stage airflow data; sort the first-stage airflow data of each area according to the position of the test point, and obtain the first airflow sorting result;

[0015] Divide the area according to the layout structure of the nozzle, set a collection point in each area, obtain the spray pressure of the collection point, and obtain the first-stage spray data; sort the first-stage spray data of each area according to the position of the collection point to obtain the first spray sorting result;

[0016] The first airflow sorting result and the first spray sorting result are combined in time sequence, and first process parameter information is obtained according to the result of the time sequence combination.

[0017] In one embodiment of the present invention, the air flow data and the spray data in the particle growth stage are measured to obtain the second process parameter information, including:

[0018] In the particle growth stage, the air flow velocity and air flow temperature of the test point and the ambient humidity in the granulation chamber are collected to obtain the second-stage air flow data; the second-stage air flow data of each area are sorted according to the position of the test point to obtain the second air flow sorting result;

[0019] In the particle growth stage, the spray pressure of the collection point is obtained to obtain the second-stage spray data; the second-stage spray data of each area is sorted according to the position of the collection point to obtain the second spray sorting result;

[0020] The second airflow sorting result and the second spray sorting result are combined in time sequence, and second process parameter information is obtained according to the result of the time sequence combination.

[0021] In one embodiment of the present invention, the first process parameter information and the second process parameter information are analyzed to obtain a process parameter analysis result, including:

[0022] Calculating the airflow parameter gradient between adjacent test points according to the first airflow sorting result and the second airflow sorting result;

[0023] Obtaining an airflow parameter gradient result set according to each of the airflow parameter gradients;

[0024] A process parameter analysis result is obtained according to the first airflow sorting result, the second airflow sorting result and the airflow parameter gradient result set.

[0025] In one embodiment of the present invention, the first process parameter information and the second process parameter information are analyzed to obtain a process parameter analysis result, including:

[0026] Calculating the spray parameter gradient between adjacent collection points according to the first spray sorting result and the second spray sorting result;

[0027] Obtaining a spray parameter gradient result set according to each of the spray parameter gradients;

[0028] A process parameter analysis result is obtained according to the first spray sequencing result, the second spray sequencing result and the spray parameter gradient result set.

[0029] In one embodiment of the present invention, the parameter gradient is calculated according to the following method:

[0030]

[0031] Among them, G1 represents the airflow parameter gradient; A1 and A2 are the airflow data of two adjacent test points respectively; D1 is the distance between two adjacent test points; G2 represents the spray pressure gradient; S1 and S2 are the spray pressures of two adjacent collection points respectively; D2 is the distance between two adjacent collection points.

[0032] In one embodiment of the present invention, the dissolution characteristics of the prepared granules are determined to obtain the dissolution information of the prepared granules, including:

[0033] The produced particles are sorted; the sorting includes sorting according to the particle size distribution, moisture content and component content of the produced particles, and the produced particles that meet the three conditions are sorted as qualified produced particles;

[0034] The qualified particles are placed in a dissolution apparatus, a plurality of sampling time points are set according to the release characteristics of the qualified particles, and the drug concentration value at each sampling time point is measured;

[0035] According to the drug concentration values ​​measured at each sampling time point, a dissolution curve is drawn with time as the horizontal axis and the percentage of drug dissolution as the vertical axis;

[0036] According to the shape and data of the dissolution curve, the dissolution information of the prepared particles is obtained.

[0037] In one embodiment of the present invention, determining the dissolution characteristics of the prepared granules to obtain the dissolution information of the prepared granules further includes:

[0038] Obtaining drug concentration values ​​of a plurality of the qualified sorted granules of the same batch at various sampling time points, and calculating the dissolution percentages according to the drug concentration values, to obtain a dissolution information set of the qualified sorted granules;

[0039] Extracting the first produced particle dissolution information from the dissolution information set of the qualified sorted produced particles, and performing similarity factor f2 calculation on the first produced particle dissolution information and the reference produced particle dissolution information;

[0040] If the similarity factor f2 is greater than or equal to 90 and less than 100, the dissolution information of the first produced particles is added to the target produced particle set;

[0041] Draw the dissolution curve according to the data in the target produced particle set;

[0042] Among them, the calculation formula of the similarity factor f2 is as follows:

[0043]

[0044] Where n represents the number of sampling time points; i represents the sequence number of the sampling time point; R i represents the dissolution percentage of the reference prepared granules at the i-th sampling time point; T i It represents the dissolution percentage of the first prepared granules at the i-th sampling time point.

[0045] In one embodiment of the present invention, obtaining the process adjustment parameters includes:

[0046] Determining whether the dissolution information of the produced particles meets a preset target value;

[0047] When the dissolution information of the produced particles does not meet the target value, parameter control and adjustment are performed on the material stage, the spraying and liquid addition stage, and the particle growth stage through the process optimization model;

[0048] The process adjustment parameters are obtained according to the parameter control adjustment.

[0049] In the second aspect, in order to solve the above technical problems, the present invention also provides a drug granulation process optimization system based on artificial intelligence, the drug granulation includes a material stage, a spraying liquid addition stage and a particle growth stage, and the system includes:

[0050] A material attribute data acquisition module, which is used to measure the material attribute data of the material stage and obtain material attribute parameter information;

[0051] A process data acquisition module, which is used to measure the airflow data and the spray data in the spray liquid addition stage to obtain the first process parameter information; and to measure the airflow data and the spray data in the particle growth stage to obtain the second process parameter information;

[0052] A parameter analysis module, which obtains a process parameter analysis result by analyzing the first process parameter information and the second process parameter information;

[0053] A dissolution information collection module for produced particles; it is used to measure the dissolution characteristics of the produced particles and obtain the dissolution information of the produced particles;

[0054] A parameter adaptive adjustment module is used to input the material attribute parameter information, the process parameter analysis results and the dissolution information of the produced particles into a process optimization model to obtain process adjustment parameters; and adaptively adjust the parameters of the material stage, the spray liquid addition stage and the particle growth stage according to the process adjustment parameters.

[0055] The above technical solution of the present invention has the following beneficial effects compared with the prior art:

[0056] The artificial intelligence-based drug granulation process optimization method and system described in the present invention effectively solve the quality problem of produced granules caused by deviations in dissolution characteristics, and ensure that granules produced from different batches of powder raw materials and materials of the same model have consistency and stability in dissolution characteristics. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to make the content of the present invention more clearly understood, the present invention is further described in detail below according to specific embodiments of the present invention in conjunction with the accompanying drawings, wherein:

[0058] Figure 1 It is a flow chart of a drug granulation process optimization method based on artificial intelligence in a preferred embodiment of the present invention;

[0059] Figure 2 A flowchart of obtaining process parameter analysis results in a preferred embodiment of the present invention;

[0060] Figure 3 A flow chart of obtaining dissolution information of produced particles in a preferred embodiment of the present invention;

[0061] Figure 4 A flow chart of obtaining process adjustment parameters in a preferred embodiment of the present invention;

[0062] Figure 5 The figure is a structural block diagram of a drug granulation process optimization system based on artificial intelligence in a preferred embodiment of the present invention. DETAILED DESCRIPTION

[0063] The present invention is further described below in conjunction with the accompanying drawings and specific embodiments so that those skilled in the art can better understand the present invention and implement it, but the embodiments are not intended to limit the present invention.

[0064] The following should be explained:

[0065] The fluidized bed wet granulation process is divided into the following stages:

[0066] Material stage: The adhesive material is configured through the material tank. The material density, material viscosity, material water content and material / auxiliary material ratio are configured before use. The configured material is used in the spray liquid addition stage and the particle growth stage.

[0067] Fluidization stage: Use the airflow from bottom to top to make the drug powder raw materials suspended and fluidized.

[0068] Spraying and liquid addition stage (also called the mother core formation stage): After the powder is in a fluidized state, the binder liquid is atomized through a nozzle and sprayed onto the fluidized bed. The droplets come into contact with the powder and form a particle core around it. This process is the initial stage of granulation and forms a preliminary particle core.

[0069] Particle growth stage (also called parent core growth stage): As the spraying continues, new droplets fall on the existing particle cores. Through the bridging effect of the adhesive, the particles combine with each other and gradually grow into larger particles.

[0070] Drying stage: After the particle growth is completed, it is dried until the particle size meets the requirements.

[0071] Embodiment 1

[0072] Reference Figure 1 As shown, the embodiment of the present invention discloses a method for optimizing a drug granulation process based on artificial intelligence, the method comprising:

[0073] S100, measuring the material property data at the material stage to obtain material property parameter information; wherein the material properties include material density, material viscosity, material moisture content, and the ratio of material to auxiliary materials. At the beginning of the granulation process, the material property data is obtained by measuring various properties of the material.

[0074] S200, measuring the air flow data and the spray data in the spray liquid addition stage to obtain first process parameter information;

[0075] S300, measuring the airflow data and the spray data in the particle growth stage to obtain second process parameter information;

[0076] In specific application scenarios, air flow data include air flow velocity and air flow temperature; spray data include spray pressure. The above air flow data and spray data have an impact on the dissolution characteristics of the produced particles, and their respective impact paths include:

[0077] The air velocity determines the movement state of the powder particles in the fluidized bed; a higher air velocity can better disperse the particles, avoid local accumulation, and ensure uniform distribution of the binder. However, too high an air velocity may cause fine powder to be taken out of the system, reducing the number of effective particles; while a lower air velocity may cause poor particle flowability, increase the risk of agglomeration, and affect the consistency of dissolution.

[0078] The air flow temperature directly determines the rate of solvent evaporation; for example, a higher air flow temperature can accelerate solvent evaporation and cause the particles to solidify quickly, which may form a denser particle structure and slow down the dissolution rate; but too high a temperature may cause the particle surface to harden rapidly and even cause the degradation of heat-sensitive drug components; while a lower temperature may prolong the drying time, making it difficult to completely remove the moisture inside the particles, which may lead to a loose particle structure and accelerate the dissolution rate.

[0079] Spray pressure is directly related to the spraying method of the binder solution, the particle formation mechanism, and the physical properties of the final particles. Among them, (1) higher spray pressures usually produce finer droplets. For example, smaller droplets can be more evenly distributed throughout the powder bed and can more easily penetrate between powder particles, thereby promoting better bonding, which can lead to a more compact and uniform particle structure, which may increase particle density and affect dissolution rate. In contrast, larger droplets produced by lower pressures may lead to uneven bonding, forming larger but loosely structured particles, which may have a faster dissolution rate. (2) Spray pressure also indirectly determines the spray rate, that is, the amount of liquid sprayed per unit time. For example, under high spray pressure, if the spray rate is too fast, it may lead to excessive local humidity, causing the particles to quickly solidify or form a shell, limiting the dissolution of the internal substances; while slower spray rates allow more drying time, which helps to form more uniform particles and is conducive to obtaining ideal dissolution behavior. (3) Spray pressure affects the way droplets come into contact with powder, which in turn affects the growth pattern of particles. For example, appropriate spray pressure helps control the growth process of particles and avoid excessive growth or agglomeration. Appropriate particle size is crucial to ensure good dissolution performance. Particles that are too large or too small may not be conducive to the effective release of drugs.

[0080] By measuring the air flow velocity, air flow temperature and spray pressure in the spray liquid addition stage and the particle growth stage online, the particle growth during the granulation process can be understood. By obtaining the first process parameter information and the second process parameter information, the granulation process can be analyzed in detail, and the air flow velocity, air flow temperature and spray pressure can be adjusted and optimized accordingly to determine the optimal process parameters to achieve better dissolution characteristics.

[0081] S400, analyzing the first process parameter information and the second process parameter information to obtain a process parameter analysis result;

[0082] In specific application scenarios, through comprehensive analysis of the air flow velocity, air flow temperature and spray pressure distribution information in the spray liquid addition stage and the particle growth stage of granulation, the air flow velocity distribution law, air flow temperature distribution law and spray pressure distribution law in different stages can be mastered. The distribution law of air flow velocity and air flow temperature is analyzed through the time axis and the structure of the air flow sieve plate to analyze the change of process parameters, and the relationship between the dissolution characteristics and the changes of air flow velocity and air flow temperature is understood. The distribution law of spray pressure is analyzed through the time axis and the layout structure of the nozzle to analyze the change of spray pressure, and the relationship between the dissolution characteristics and the spray pressure is understood.

[0083] S500, measuring the dissolution characteristics of the produced granules to obtain dissolution information of the produced granules;

[0084] In specific application scenarios, the produced granules refer to the drug granules obtained after the drying stage; the dissolution characteristics of drug granules refer to the speed and degree of dissolution of drug granules in a specific medium, which is one of the key indicators for evaluating the quality and bioavailability of drug preparations. Determining the dissolution characteristics of produced granules is a key means to determine whether the granulation is qualified.

[0085] S600, inputting the material attribute parameter information, the process parameter analysis results and the dissolution information of the produced particles into a process optimization model to obtain process adjustment parameters;

[0086] In a specific application scenario, the material attribute parameter information, the process parameter analysis results and the dissolution information of the produced particles are input into a process optimization model. The model should be based on the characteristics and requirements of the granulation process to establish an adaptive adjustment model for the material stage, the liquid addition spray stage and the particle growth stage. For example, it can be a model established based on a neural network model, a genetic algorithm and other methods. According to the model calculation and analysis results, the optimal process parameters and material attribute parameters are obtained. The process parameters include air flow velocity, air flow temperature, spray pressure, humidity in the granulation chamber, etc. The material attribute parameters include material viscosity, material density, material moisture content and material / auxiliary material ratio, etc.

[0087] S700, adaptively adjusting the parameters of any of the three stages: the material stage, the spraying and liquid addition stage, and the particle growth stage according to the process adjustment parameters.

[0088] In specific application scenarios, according to the process adjustment parameters and material adjustment parameters obtained by the process optimization model, the parameters of the material stage, liquid addition spray stage and particle growth stage are automatically adjusted during the actual granulation process. The particles produced by the same batch of powder raw materials and materials have similar dissolution characteristics. By analyzing the dissolution characteristics of the current particles and obtaining the corresponding process parameters and material attribute parameter changes, the process parameters and material attribute parameters of the next batch of granulation can be adjusted. By measuring the process parameters, material attribute parameters and particle dissolution characteristics, they are input into the process optimization model to obtain the best process adjustment parameters, and then the parameters of each stage are automatically adjusted according to these parameters.

[0089] The artificial intelligence-based drug granulation process optimization method described in the present invention effectively solves the quality problem of produced granules caused by deviations in dissolution characteristics, and ensures that granules produced from different batches of powder raw materials and materials of the same model have consistency and stability in dissolution characteristics.

[0090] Furthermore, the air flow data and the spray data in the spray liquid addition stage are measured to obtain the first process parameter information, including:

[0091] Divide the area according to the structure of the airflow screen plate, set a test point in each area, collect the airflow velocity and airflow temperature of the test point, and obtain the first-stage airflow data; sort the first-stage airflow data of each area according to the position of the test point, and obtain the first airflow sorting result;

[0092] In specific application scenarios, the area is divided according to the structure of the airflow screen plate, and each area represents a specific airflow speed range and airflow temperature range. For example, according to the position of the sieve holes on the airflow screen plate, it can be divided according to a single sieve hole, that is, one sieve hole corresponds to one position; or multiple adjacent sieve holes can correspond to one position, and the airflow speed and airflow temperature corresponding to each position can be regarded as an area, recorded as the first area to measure the airflow parameters, and the positions of the test points corresponding to the first area measured airflow parameters of each area can be sorted, and the sorting can be sorted in order according to the positions of the test points to determine the airflow parameters at each position and the changes in the airflow parameters of the adjacent areas, and obtain the first airflow sorting result.

[0093] Divide the area according to the layout structure of the nozzle, set a collection point in each area, obtain the spray pressure of the collection point, and obtain the first-stage spray data; sort the first-stage spray data of each area according to the position of the collection point to obtain the first spray sorting result;

[0094] In specific application scenarios, the area division is performed according to the layout structure of the nozzles, and each area represents a specific spray pressure range. For example, according to the position and number of nozzles, one nozzle can correspond to one position, and the spray pressure corresponding to each position can be regarded as an area, which is recorded as the first area measured spray pressure. The positions of the collection points corresponding to the first area measured spray pressure of each area are sorted, and the sorting can be done in sequence according to the positions of the collection points to determine the spray pressure at each position and the spray pressure changes in the adjacent areas, and obtain the first spray sorting result.

[0095] The first airflow sorting result and the first spray sorting result are combined in time sequence, and first process parameter information is obtained according to the result of the time sequence combination.

[0096] In a specific application scenario, the first process parameter information can be in the form of a matrix. When constructing the matrix, it is obtained based on the time point, the divided area of ​​the airflow screen plate and the divided area of ​​the nozzle. The elements in the matrix include the airflow condition of an area at a certain time and the spray pressure condition of an area at a certain time.

[0097] Further, the airflow data and spray data in the particle growth stage are measured to obtain second process parameter information, including:

[0098] In the particle growth stage, the air flow velocity and air flow temperature of the test point and the ambient humidity in the granulation chamber are collected to obtain the second-stage air flow data; the second-stage air flow data of each area are sorted according to the position of the test point to obtain the second air flow sorting result;

[0099] In the particle growth stage, the spray pressure of the collection point is obtained to obtain the second-stage spray data; the second-stage spray data of each area is sorted according to the position of the collection point to obtain the second spray sorting result;

[0100] The second airflow sorting result and the second spray sorting result are combined in time sequence, and second process parameter information is obtained according to the result of the time sequence combination.

[0101] In a specific application scenario, the second airflow sorting result adopts the same division and calculation method as the first airflow sorting result; and the second spray sorting result adopts the same division and calculation method as the first spray sorting result, and finally obtains the second process parameter information.

[0102] Further, refer to Figure 2 As shown, the first process parameter information and the second process parameter information are analyzed to obtain process parameter analysis results, including:

[0103] Calculating the airflow parameter gradient between adjacent test points according to the first airflow sorting result and the second airflow sorting result;

[0104] Obtaining an airflow parameter gradient result set according to each of the airflow parameter gradients;

[0105] A process parameter analysis result is obtained according to the first airflow sorting result, the second airflow sorting result and the airflow parameter gradient result set.

[0106] In the specific application scenario, the first airflow sorting result of the spray and liquid addition stage and the second airflow sorting result of the particle growth stage are matched according to the same position to obtain the third airflow sorting result; the airflow parameter gradient between adjacent test points is calculated, where the airflow parameter gradient includes the airflow velocity gradient and the airflow temperature gradient. The airflow velocity gradient refers to the ratio of the airflow velocity change to the distance unit; the airflow temperature gradient refers to the ratio of the airflow temperature change to the distance unit; the calculated airflow parameter gradients are summarized according to the position of each test point to form the airflow parameter analysis result.

[0107] In one embodiment of the present invention, the first process parameter information and the second process parameter information are analyzed to obtain a process parameter analysis result, including:

[0108] Calculating the spray parameter gradient between adjacent collection points according to the first spray sorting result and the second spray sorting result;

[0109] Obtaining a spray parameter gradient result set according to each of the spray parameter gradients;

[0110] A process parameter analysis result is obtained according to the first spray sequencing result, the second spray sequencing result and the spray parameter gradient result set.

[0111] In a specific application scenario, the first spray sorting result of the spray dosing stage and the second spray sorting result of the particle growth stage are matched at the same position to obtain a third spray sorting result; the spray parameter gradient between adjacent collection points is calculated, and the spray parameter gradient refers to the ratio of the spray pressure change unit to the distance unit; the calculated spray parameter gradient is summarized according to the position of each collection point to form a spray parameter analysis result.

[0112] The process parameter analysis results are obtained based on the air flow parameter analysis results and the spray parameter analysis results.

[0113] Furthermore, the parameter gradient is calculated as follows:

[0114]

[0115] Among them, when calculating the air flow velocity gradient: G1 represents the air flow velocity gradient, A1 and A2 are the air flow velocities of two adjacent test points respectively; when calculating the air flow temperature gradient, G1 represents the air flow temperature gradient, A1 and A2 are the air flow temperatures of two adjacent test points respectively; D1 is the distance between two adjacent test points; G2 represents the spray pressure gradient; S1 and S2 are the spray pressures of two adjacent collection points respectively; D2 is the distance between two adjacent collection points.

[0116] In specific application scenarios, the size of the airflow velocity gradient is related to the rate of change of the airflow velocity. A larger airflow velocity gradient indicates a larger change in airflow velocity per unit length or unit distance. Similarly, the size of the airflow temperature gradient is related to the rate of change of the airflow temperature. A larger airflow temperature gradient indicates a larger change in airflow temperature per unit length or unit distance. The size of the spray pressure gradient is related to the rate of change of the spray pressure. Generally speaking, the distance between adjacent test points and the distance between adjacent collection points are fixed values, that is, D1 and D2 are usually fixed values. If the airflow velocity / airflow temperature of adjacent test points changes greatly, the airflow velocity / airflow temperature gradient will be larger; if the spray pressure of adjacent collection points changes greatly, the spray pressure gradient will be larger; otherwise, it will be smaller.

[0117] Further, refer to Figure 3 As shown, the dissolution characteristics of the prepared granules are measured to obtain dissolution information of the prepared granules, including:

[0118] The produced granules are sorted; the sorting includes sorting according to the particle size distribution, water content and component content of the produced granules, and the produced granules that meet the three conditions are sorted as qualified produced granules; through strict sorting of the particle size distribution, water content and component content, it is ensured that the produced granules used for the dissolution test have a high degree of consistency and representativeness, thereby reducing the error caused by sample differences.

[0119] The qualified sorted granules are placed in a dissolution apparatus, and a plurality of sampling time points are set according to the release characteristics of the qualified sorted granules, such as 5 minutes, 10 minutes, 15 minutes, 30 minutes, 60 minutes, etc.; and the drug concentration value at each of the sampling time points is measured;

[0120] According to the drug concentration values ​​measured at each sampling time point, a dissolution curve is drawn with time as the horizontal axis and the percentage of drug dissolution as the vertical axis; wherein the dissolution percentage = (dissolution amount / labeled amount) × 100%;

[0121] According to the shape and data of the dissolution curve, the dissolution characteristics of the prepared particles are evaluated, including the initial dissolution rate, the maximum dissolution percentage, the time required to reach a specific dissolution percentage, etc., to obtain the dissolution information of the prepared particles.

[0122] Further, determining the dissolution characteristics of the prepared granules to obtain dissolution information of the prepared granules also includes:

[0123] Obtaining drug concentration values ​​of a plurality of the qualified sorted granules of the same batch at various sampling time points, and calculating the dissolution percentages according to the drug concentration values, to obtain a dissolution information set of the qualified sorted granules;

[0124] Extracting the first produced particle dissolution information from the dissolution information set of the qualified sorted produced particles, and performing similarity factor f2 calculation on the first produced particle dissolution information and the reference produced particle dissolution information;

[0125] Specifically, any one of the qualified produced particles in the sorting is extracted as the first produced particle, and the dissolution information of the first produced particle is extracted to obtain the dissolution information of the first produced particle; the reference produced particle refers to the particle whose dissolution characteristics meet the requirements, and whether the dissolution information of the first produced particle is valid is judged by comparing the similarity of the dissolution characteristics between the first produced particle and the reference produced particle.

[0126] If the similarity factor f2 is greater than or equal to 90 and less than 100, the dissolution information of the first produced particles is added to the target produced particle set; specifically, the similarity factor f2 is used to intuitively evaluate the similarity of the dissolution characteristics between the first produced particles and the reference produced particles. The closer the similarity factor is to 100, the more similar the dissolution characteristic curves of the first produced particles and the reference produced particles are.

[0127] The dissolution curve is drawn according to the data in the target produced particle set; specifically, the similarity factor f2 between the multiple sorted qualified produced particles of the same batch and the reference produced particles is calculated respectively, and the dissolution curve is drawn only according to the dissolution characteristic data of the sorted qualified produced particles that meet the requirements of the similarity factor, thereby greatly improving the accuracy of the dissolution characteristic measurement.

[0128] Among them, the calculation formula of the similarity factor f2 is as follows:

[0129]

[0130] Where n represents the number of sampling time points; i represents the sequence number of the sampling time point; R i represents the dissolution percentage of the reference prepared granules at the i-th sampling time point; T i It represents the dissolution percentage of the first prepared granules at the i-th sampling time point.

[0131] Further, refer to Figure 4 As shown, the process adjustment parameters include:

[0132] Determining whether the dissolution information of the produced particles meets a preset target value;

[0133] When the dissolution information of the produced particles does not meet the target value, parameter control and adjustment are performed on the material stage, the spraying and liquid addition stage, and the particle growth stage through the process optimization model;

[0134] The process adjustment parameters are obtained according to the parameter control adjustment.

[0135] In specific application scenarios, first determine whether the dissolution information of the produced particles meets the preset target value. If it does, it means that the quality of the produced particles meets the requirements; if not, the dissolution information of the batch of produced particles is imported into the process optimization model, and the dissolution characteristics of the produced particles are made to meet the preset target value by adjusting the air flow parameters, spray pressure, and material parameters. Generally speaking, the process optimization model is implemented through adaptive optimization methods such as machine learning and genetic algorithms to obtain the adjusted process parameters.

[0136] Embodiment 2

[0137] Based on the same inventive concept as the artificial intelligence-based drug granulation process optimization method in the aforementioned embodiment, the present invention also provides an artificial intelligence-based drug granulation process optimization system, wherein the drug granulation includes a material stage, a spraying liquid addition stage, and a particle growth stage. Figure 5 As shown, the system comprises:

[0138] A material attribute data acquisition module, which is used to measure the material attribute data of the material stage and obtain material attribute parameter information;

[0139] A process data acquisition module, which is used to measure the airflow data and the spray data in the spray liquid addition stage to obtain the first process parameter information; and to measure the airflow data and the spray data in the particle growth stage to obtain the second process parameter information;

[0140] A parameter analysis module, which obtains a process parameter analysis result by analyzing the first process parameter information and the second process parameter information;

[0141] A dissolution information collection module for produced particles; it is used to measure the dissolution characteristics of the produced particles and obtain the dissolution information of the produced particles;

[0142] A parameter adaptive adjustment module is used to input the material attribute parameter information, the process parameter analysis results and the dissolution information of the produced particles into a process optimization model to obtain process adjustment parameters; and adaptively adjust the parameters of the material stage, the spray liquid addition stage and the particle growth stage according to the process adjustment parameters.

[0143] The process optimization system for wet drug granulation of the embodiment of the present invention is used to execute the drug granulation process optimization method based on artificial intelligence. The technical effects that can be achieved are as described in the above embodiment and will not be repeated here.

[0144] In summary, the artificial intelligence-based drug granulation process optimization method and system described in the present invention effectively solve the quality problems of the produced granules caused by the deviation of the dissolution characteristics, and ensure that the granules produced from different batches of powder raw materials and materials of the same model have consistency and stability in dissolution characteristics.

[0145] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0146] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0147] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0148] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0149] Obviously, the above embodiments are merely examples for clear explanation and are not intended to limit the implementation methods. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the implementation methods here. The obvious changes or modifications derived from these are still within the protection scope of the invention.

Claims

1. A drug granulation process optimization method based on artificial intelligence, drug granulation includes a material stage, a spraying liquid addition stage and a particle growth stage, characterized in that: The method comprises, Determine the material attribute data at the material stage to obtain material attribute parameter information; Determine the air flow data and the spray data in the spray liquid addition stage to obtain first process parameter information; measuring the air flow data and the spray data in the particle growth stage to obtain second process parameter information; Analyze the first process parameter information and the second process parameter information to obtain a process parameter analysis result; Determine the dissolution characteristics of the prepared granules to obtain dissolution information of the prepared granules; Inputting the material attribute parameter information, the process parameter analysis results and the dissolution information of the produced particles into a process optimization model to obtain process adjustment parameters; According to the process adjustment parameters, the parameters of any of the three stages, namely, the material stage, the spraying and liquid addition stage, and the particle growth stage, are adaptively adjusted.

2. The method for optimizing drug granulation process based on artificial intelligence according to claim 1, characterized in that: Determine the air flow data and spray data of the spray liquid addition stage to obtain first process parameter information, including: Divide the area according to the structure of the airflow screen plate, set a test point in each area, collect the airflow velocity and airflow temperature of the test point, and obtain the first-stage airflow data; sort the first-stage airflow data of each area according to the position of the test point, and obtain the first airflow sorting result; Divide the area according to the layout structure of the nozzle, set a collection point in each area, obtain the spray pressure of the collection point, and obtain the first-stage spray data; sort the first-stage spray data of each area according to the position of the collection point to obtain the first spray sorting result; The first airflow sorting result and the first spray sorting result are combined in time sequence, and first process parameter information is obtained according to the result of the time sequence combination.

3. The method for optimizing drug granulation process based on artificial intelligence according to claim 2, characterized in that: Determine the airflow data and spray data during the particle growth stage to obtain second process parameter information, including: In the particle growth stage, the air flow velocity and air flow temperature of the test point and the ambient humidity in the granulation chamber are collected to obtain the second-stage air flow data; the second-stage air flow data of each area are sorted according to the position of the test point to obtain the second air flow sorting result; In the particle growth stage, the spray pressure of the collection point is obtained to obtain the second-stage spray data; the second-stage spray data of each area is sorted according to the position of the collection point to obtain the second spray sorting result; The second airflow sorting result and the second spray sorting result are combined in time sequence, and second process parameter information is obtained according to the result of the time sequence combination.

4. The method for optimizing drug granulation process based on artificial intelligence according to claim 3, characterized in that: Analyze the first process parameter information and the second process parameter information to obtain a process parameter analysis result, including: Calculating the airflow parameter gradient between adjacent test points according to the first airflow sorting result and the second airflow sorting result; Obtaining an airflow parameter gradient result set according to each of the airflow parameter gradients; A process parameter analysis result is obtained according to the first airflow sorting result, the second airflow sorting result and the airflow parameter gradient result set.

5. The method for optimizing drug granulation process based on artificial intelligence according to claim 3 or 4, characterized in that: Analyze the first process parameter information and the second process parameter information to obtain a process parameter analysis result, including: Calculating the spray parameter gradient between adjacent collection points according to the first spray sorting result and the second spray sorting result; Obtaining a spray parameter gradient result set according to each of the spray parameter gradients; A process parameter analysis result is obtained according to the first spray sequencing result, the second spray sequencing result and the spray parameter gradient result set.

6. The method for optimizing drug granulation process based on artificial intelligence according to claim 4, characterized in that: The parameter gradients are calculated as follows: Among them, G1 represents the airflow parameter gradient; A1 and A2 are the airflow data of two adjacent test points respectively; D1 is the distance between two adjacent test points; G2 represents the spray pressure gradient; S1 and S2 are the spray pressures of two adjacent collection points respectively; D2 is the distance between two adjacent collection points.

7. The method for optimizing drug granulation process based on artificial intelligence according to claim 1, characterized in that: Determine the dissolution characteristics of the prepared granules and obtain dissolution information of the prepared granules, including: The produced particles are sorted; the sorting includes sorting according to the particle size distribution, moisture content and component content of the produced particles, and the produced particles that meet the three conditions are sorted as qualified produced particles; The qualified particles are placed in a dissolution apparatus, a plurality of sampling time points are set according to the release characteristics of the qualified particles, and the drug concentration value at each sampling time point is measured; According to the drug concentration values ​​measured at each sampling time point, a dissolution curve is drawn with time as the horizontal axis and the percentage of drug dissolution as the vertical axis; According to the shape and data of the dissolution curve, the dissolution information of the prepared particles is obtained.

8. The method for optimizing drug granulation process based on artificial intelligence according to claim 7, characterized in that: Determining the dissolution characteristics of the prepared granules to obtain dissolution information of the prepared granules, further comprising: Obtaining drug concentration values ​​of a plurality of the qualified sorted granules of the same batch at various sampling time points, and calculating the dissolution percentages according to the drug concentration values, to obtain a dissolution information set of the qualified sorted granules; Extracting the first produced particle dissolution information from the dissolution information set of the qualified sorted produced particles, and performing similarity factor f2 calculation on the first produced particle dissolution information and the reference produced particle dissolution information; If the similarity factor f2 is greater than or equal to 90 and less than 100, the dissolution information of the first produced particles is added to the target produced particle set; Draw the dissolution curve according to the data in the target produced particle set; Among them, the calculation formula of the similarity factor f2 is as follows: Where n represents the number of sampling time points; i represents the sequence number of the sampling time point; R i represents the dissolution percentage of the reference prepared granules at the i-th sampling time point; T i It represents the dissolution percentage of the first prepared granules at the i-th sampling time point.

9. The method for optimizing drug granulation process based on artificial intelligence according to claim 1, characterized in that: Obtaining process adjustment parameters includes: Determining whether the dissolution information of the produced particles meets a preset target value; When the dissolution information of the produced particles does not meet the target value, parameter control and adjustment are performed on the material stage, the spraying and liquid addition stage, and the particle growth stage through the process optimization model; The process adjustment parameters are obtained according to the parameter control adjustment.

10. A drug granulation process optimization system based on artificial intelligence, drug granulation includes a material stage, a spraying liquid addition stage and a particle growth stage, characterized in that: The system comprises, A material attribute data acquisition module, which is used to measure the material attribute data of the material stage and obtain material attribute parameter information; A process data acquisition module, which is used to measure the airflow data and the spray data in the spray liquid addition stage to obtain the first process parameter information; and to measure the airflow data and the spray data in the particle growth stage to obtain the second process parameter information; A parameter analysis module, which obtains a process parameter analysis result by analyzing the first process parameter information and the second process parameter information; Produce a particle dissolution information collection module; It is used to measure the dissolution characteristics of the prepared granules and obtain the dissolution information of the prepared granules; A parameter adaptive adjustment module is used to input the material attribute parameter information, the process parameter analysis results and the dissolution information of the produced particles into a process optimization model to obtain process adjustment parameters; and adaptively adjust the parameters of the material stage, the spray liquid addition stage and the particle growth stage according to the process adjustment parameters.

Images