Intelligent analysis method for bulk drug crystal form transformation risk

Abstract

The invention particularly relates to the technical field of material analysis, and discloses an intelligent analysis method for bulk drug crystal form transformation risk, which comprises the following steps: S1, data acquisition and preprocessing; s2, data comprehensive analysis and evaluation; s3, crystal form transformation risk judgment; s4, generating an adjustment control instruction; s5, issuing and executing the regulation and control instruction; s6, feedback optimization closed-loop control is carried out, crystal form transformation related parameters are collected in real time through an online detection terminal, a crystal form transformation risk assessment model is constructed through a data processing terminal, historical production data are integrated, and a risk assessment result is obtained. Performing coupling calculation on the three-dimensional crystal form characteristic risk index, the crystal form environment risk index and the regulation and control process risk index to obtain a crystal form conversion comprehensive risk index, judging a risk level based on a preset risk threshold value and identifying a key risk influence factor, and generating a regulation and control instruction according to a preset regulation and control rule, and regulation effect data are monitored and fed back in real time to optimize model parameters and regulation rules, so that the accuracy of risk analysis is remarkably improved, and the dynamic adaptability of crystal form judgment is enhanced.

Description

Technical Field

[0001] This invention relates to the field of materials analysis technology, and more specifically, to an intelligent analysis method for the risk of crystal form transformation of active pharmaceutical ingredients. Background Technology

[0002] Different crystal forms of active pharmaceutical ingredients (APIs) may have different physicochemical properties and bioavailability, affecting efficacy. The crystal form of APIs directly affects key quality attributes such as dissolution rate, solubility, stability, and bioavailability. As drug regulatory authorities impose increasingly stringent requirements on drug quality control, the control of this key quality attribute of API crystal form is becoming increasingly important. With the development of computer technology, the use of computer simulation to assist in drug crystal form screening and risk analysis is becoming a trend.

[0003] Existing methods for analyzing the risk of polymorph transformation in active pharmaceutical ingredients (APIs) include polymorph screening and characterization, polymorph transformation mechanism research, polymorph transformation risk identification, polymorph transformation risk assessment, and control strategy formulation, which can specifically reduce the adverse effects of polymorph transformation on the quality, safety, and efficacy of APIs.

[0004] However, it still has some drawbacks in practical use. First, the analysis efficiency is low. The existing methods for analyzing the risk of polymorph transformation of active pharmaceutical ingredients rely heavily on manual operation and experience judgment in the polymorph screening and transformation mechanism research stages. Each stage is relatively independent, resulting in low data transmission and integration efficiency, making it difficult to cope with complex and multivariate scenarios. Second, the accuracy of risk prediction is limited. Existing methods for analyzing the risk of crystal transformation of active pharmaceutical ingredients are mostly based on small sample experiments or qualitative analysis. They lack in-depth mining and analysis of massive amounts of data and cannot quantify the risk of crystal transformation under complex conditions, resulting in prediction bias and a lack of data-driven precision optimization. Third, the dynamic adaptability is insufficient. Existing methods for analyzing the risk of crystal form transformation in active pharmaceutical ingredients (APIs) are mostly offline detection methods. They cannot capture dynamic changes in the production and storage process in real time, making it difficult to cope with fluctuations in the API production process and to adjust control strategies in a timely manner. They lack dynamic adaptability and real-time performance. Summary of the Invention

[0005] In view of this, embodiments of the present invention provide an intelligent analysis method for the crystal form transformation risk of active pharmaceutical ingredients. The method collects crystal form transformation-related parameters in real time through an online detection terminal, constructs a crystal form transformation risk assessment model through a data processing terminal, determines the risk level based on a preset risk threshold and identifies key risk influencing factors, generates control instructions according to preset control rules, and monitors and provides feedback on control effect data in real time to optimize model parameters and control rules. This method effectively solves the problems of low analysis efficiency, limited risk prediction accuracy, and insufficient dynamic adaptability mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: an intelligent analysis method for the crystal form transformation risk of a pharmaceutical active pharmaceutical ingredient, comprising an online detection terminal, a data processing terminal, and an intelligent control terminal, wherein each terminal is connected via a real-time communication protocol, and the steps are as follows: S1: Data Acquisition and Preprocessing: Crystal form transformation related parameters are acquired in real time through an online detection terminal, and then transmitted to the data processing terminal after data preprocessing; S2: Comprehensive Data Analysis and Evaluation: The data processing terminal constructs a crystal form transformation risk assessment model and calculates the comprehensive crystal form transformation risk index based on the crystal form transformation-related parameters. S3: Crystal form transformation risk assessment: Preset a comprehensive risk threshold for crystal form transformation, determine the risk level of crystal form transformation based on the comprehensive risk index of crystal form transformation, and identify key risk influencing factors; S4: Adjustment of control command generation: Based on preset control rules, the intelligent control terminal generates control commands for production equipment according to the risk level and key risk influencing factors; S5: Issuance and execution of control instructions: The intelligent control terminal transmits control instructions for the production equipment to the production equipment, executes the control instructions, and provides real-time feedback on the control effect data. S6: Feedback-optimized closed-loop control: The data processing terminal updates the crystal form transformation risk assessment model and control rules based on the control effect data.

[0007] The technical effects and advantages of this invention are as follows: This invention collects crystal form transformation-related parameters in real time through an online detection terminal, transmits the data to a data processing terminal, and realizes closed-loop control of data preprocessing, risk assessment, instruction generation and execution, and feedback adjustment through automated algorithms. This effectively solves the problem of low analysis efficiency caused by manual operation and significantly improves the accuracy of risk analysis. This invention constructs a crystal form transformation risk assessment model through a data processing terminal, integrates historical production data, and calculates a comprehensive crystal form transformation risk index by coupling a three-dimensional crystal form characteristic risk index, a crystal form environmental risk index, and a regulation process risk index. This avoids high deviation rates caused by single threshold judgments and improves the accuracy of risk analysis. This invention collects dynamic data in real time through an online detection terminal and performs real-time analysis in conjunction with a data processing terminal to achieve real-time capture of the crystal form transformation of active pharmaceutical ingredients (APIs) and to perform feedback optimization closed-loop control. By dynamically updating model parameters and control rules through the control effect data, it continuously adapts to process fluctuations, shortens response delay time, enhances the dynamic adaptability of crystal form judgment, and realizes real-time control and closed-loop optimization in the API production process. Attached Figure Description

[0008] Figure 1 This is a schematic diagram of the method steps of the present invention.

[0009] Figure 2 This is a schematic diagram of the overall structure of the present invention.

[0010] Figure 3 This is a schematic diagram of the crystal form transformation risk level determination steps of the present invention. Detailed Implementation

[0011] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0012] As attached Figure 1 The method shown is an intelligent analysis method for the crystal form transformation risk of a pharmaceutical active pharmaceutical ingredient, which includes an online detection terminal, a data processing terminal, and an intelligent control terminal, all of which are connected via a real-time communication protocol.

[0013] In a more specific application of the present invention, the online detection terminal is used to collect multi-dimensional raw data related to crystal form transformation in real time and continuously monitor crystal form state and influencing factors. It includes physicochemical parameter sensors, crystal form feature detection equipment, and data preprocessing unit. The physicochemical parameter sensors include environmental parameter sensors and process parameter sensors. The crystal form feature detection equipment includes an online spectrometer, an online particle size analyzer, and an online X-ray module. The online detection terminal transmits the preprocessed data to the data processing terminal via industrial Ethernet.

[0014] The data processing terminal is used to perform in-depth analysis of the raw data transmitted by the online detection terminal, realizing intelligent identification, assessment, and early warning of crystal form transformation risks. The data processing terminal includes a data storage and integration module, an intelligent algorithm engine, and a visualization interface. The data storage and integration module includes a distributed database and a data fusion unit. The intelligent algorithm engine is used to analyze and calculate the comprehensive risk index of crystal form transformation and determine whether any abnormalities have occurred. The visualization interface includes a real-time dashboard and an early warning output module. The data processing terminal sends frequency adjustment commands to the online detection terminal via industrial Ethernet and connects to the intelligent control terminal through a dedicated data interface to transmit risk assessment results.

[0015] The intelligent control terminal is used to automatically adjust production parameters based on the risk assessment results output by the data processing terminal, thereby achieving risk control over crystal form transformation. The intelligent control terminal includes a control decision unit, execution equipment, and a feedback monitoring module. The control decision unit generates control instructions based on the built-in rule engine. The execution equipment includes environmental control equipment and process parameter adjustment equipment. The feedback monitoring module collects the operating status of the execution equipment in real time and uploads the control effect data to the data processing terminal. The intelligent control terminal receives control instructions and feeds back data through the working bus, and connects to the production equipment through relays or PLCs.

[0016] For the connection methods of the aforementioned online detection terminal, data processing terminal, and intelligent control terminal, please refer to [link / reference]. Figure 2 .

[0017] The specific embodiments of the present invention include the following steps: S1: Data Acquisition and Preprocessing: Crystal form transformation related parameters are acquired in real time through an online detection terminal, and then transmitted to the data processing terminal after data preprocessing; Furthermore, the crystal form transformation-related parameters include crystal form characteristic parameters, crystal form environment parameters, and control process parameters. Among them, crystal form characteristic parameters include spectral quantification parameters, diffraction quantification parameters, and morphology quantification parameters; crystal form environment parameters include temperature and humidity quantification parameters and gas parameters; and control process parameters include material process parameters and time and energy parameters.

[0018] In this embodiment, it is necessary to specifically explain that the spectral quantification parameters include, but are not limited to, characteristic peak intensity ratio, Raman peak position shift, XRD characteristic peak area ratio, and characteristic wavelength absorbance difference; the diffraction quantification parameters include, but are not limited to, interplanar spacing deviation and polymorphic content ratio; the morphology quantification parameters include, but are not limited to, equivalent spherical diameter and crystal integrity index; the temperature and humidity quantification parameters include, but are not limited to, temperature gradient, humidity dew point, and environmental fluctuation coefficient; the gas parameters include, but are not limited to, oxygen partial pressure, carbon dioxide concentration, and gas exchange rate; the material process parameters include, but are not limited to, solution supersaturation, solvent ratio, and solid-liquid ratio; and the time-energy parameters include, but are not limited to, crystallization residence time, drying energy consumption, and cooling rate gradient.

[0019] It should be specifically noted that crystal form characteristic parameters are used to reflect the stability of the crystal form itself and can directly reflect the structural characteristics of the current crystal form of the active pharmaceutical ingredient. The greater the deviation, the easier it is to cause crystal form transformation. Crystal form environment parameters can accelerate or inhibit crystal form transformation by changing molecular motion energy or interfacial forces. Under extreme conditions, they may directly trigger transformation. Control process parameters directly affect crystal nucleation and growth during the production of active pharmaceutical ingredients. Operational deviations may lead to insufficient purity of the target crystal form or direct generation of impurity crystals.

[0020] S2: Comprehensive Data Analysis and Evaluation: The data processing terminal constructs a crystal form transformation risk assessment model and calculates the comprehensive crystal form transformation risk index based on the crystal form transformation-related parameters. Furthermore, the steps for constructing the crystal form transformation risk assessment model are as follows: A1: Parameter screening and weight allocation: Screen out key parameters related to crystal form transformation from the parameters related to crystal form transformation, and assign weights to each parameter; In this embodiment, it should be specifically explained that the key parameters related to crystal form transformation include key parameters of crystal form characteristics, key parameters of crystal form environment, and key parameters of the control process. Among them, the key parameters of crystal form characteristics are the area ratio of XRD characteristic peaks, the Raman peak position shift, and the crystal integrity index; the key parameters of crystal form environment are humidity dew point, temperature gradient, and environmental fluctuation coefficient; the key parameters of control process are solution supersaturation, cooling rate gradient, and crystallization residence time; the weight coefficients of each parameter can be 0.15, 0.15, 0.1, 0.12, 0.1, 0.08, 0.15, 0.08, and 0.07.

[0021] A2: Parameter Standardization: Using the standardization formula: , Mapping the key parameters associated with crystal form transformation to the [0,1] interval, X i S represents the key parameter associated with the i-th crystal form transformation. i X represents the standardized key parameter associated with the i-th crystal form transformation. i安全阈值 X represents the safety threshold of the key parameter associated with the i-th crystal form transformation. i危险阈值 This represents the danger threshold of the key parameter associated with the i-th crystal form transformation; In this embodiment, it needs to be specifically explained that the key parameters related to crystal form transformation are indexed as i=1, 2, ..., n, where n=9. Specifically, i=1 to 3 are key parameters of crystal form characteristics, i=4 to 6 are key parameters of the crystal form environment, and i=7 to 9 are key parameters of the control process. k S is the key parameter for the standardized crystal form characteristics. j S is the key parameter for the standardized crystal form environment. p For the key parameters of the standardized control process, assign a unique identifier to each parameter i.

[0022] It should be specifically noted that the safety threshold was obtained by selecting 50 batches of historical batch data without crystal form transformation records. After removing outliers using the box plot method, the upper limit of the 95% confidence interval for each parameter was removed. For new compounds, the phase transition enthalpy point, the critical hygroscopic point of the dynamic moisture adsorption curve, and the lower limit of the crystal form stability interval were determined by differential scanning calorimetry and in-situ XRD monitoring. The determination was made after verification by three parallel experiments. The danger threshold was obtained by collecting data from 20 batches of failed batches that had undergone crystal form transformation. The lower limit of the core interval of the parameter distribution was extracted using K-means clustering.

[0023] A3: Calculation of risk indicators by dimension: Based on the standardized parameters, the crystal form characteristic risk index FRI, the crystal form environmental risk index ERI, and the regulation process risk index PRI are calculated. In this embodiment, it is necessary to specifically explain how the standardized key parameters of the crystal form characteristics are expressed using the following formula: , The crystal form characteristic risk index FRI is calculated, where ω k The weighting coefficients represent the key parameters of each crystal form characteristic; these are obtained by applying the standardized key parameters of the crystal form environment using the formula: , The crystalline environmental risk index ERI was calculated, where ω j This represents the weighting coefficients of key environmental parameters for each crystal form; the standardized key parameters of the control process are expressed by the formula: , The risk index PRI for the regulation process is calculated, where ω p This represents the weighting coefficients of key parameters in each regulation process.

[0024] Specifically, the Crystal Form Characteristic Risk Index (FRI) is used to directly quantify the stability risk of the crystal form state of a substance itself. By capturing the intrinsic signs of crystal form transformation, it determines whether a potential transformation has occurred and uses linear weighting to directly reflect the cumulative risk of each parameter. The Crystal Form Environmental Risk Index (ERI) is used to assess the risk of external environmental factors inducing crystal form stability, reflecting the driving effect of environmental stress on crystal form transformation. It is beneficial for early warning of sudden changes caused by extreme environments and amplifies the risk contribution corresponding to high standard values ​​of parameters through an exponential function. The Control Process Risk Index (PRI) focuses on the control deviation of crystallization industry parameters on crystal form formation, quantifies the cumulative impact of process fluctuations on crystal form transformation, traces key operational nodes that may trigger crystal form transformation during production, and enhances the sensitivity of parameters to standard values ​​through a logarithmic function.

[0025] A4: Comprehensive Risk Index Coupling: Based on the crystal form characteristic risk index, crystal form environmental risk index, and control process risk index, the following formula is used: , The comprehensive risk index (CRI) for crystal form transformation was calculated. In this embodiment, it is necessary to specifically explain that the comprehensive risk index of crystal form transformation is obtained by coupling the three-dimensional indices of crystal form characteristic risk index, crystal form environmental risk index and regulation process risk index. 0.001 is used as a minimum value to avoid the denominator of the formula being 0, and multiplying by 100 maps the value of CRI to the interval [0, 300], which is convenient for subsequent risk level classification.

[0026] A5: Model Weight Correction: Verify and correct the model weights.

[0027] In this embodiment, it is necessary to specifically explain that the verification model weights require the collection of historical batch data. The comprehensive risk index of crystal form transformation calculated by the model is compared with the actual transformation results. If it is found that the explanatory power of a certain dimension sub-index for the actual transformation is insufficient, it indicates that there is a large error in the weights. Correcting the model weights requires the use of machine learning algorithms to recalculate the contribution of each sub-index to crystal form transformation and adjust the weights in each sub-index.

[0028] Furthermore, obtaining the comprehensive risk index for crystal form transformation requires setting a time window, acquiring the crystal form transformation-related parameters within that time window, importing these parameters into the crystal form transformation risk assessment model, and calculating the comprehensive risk index (FRI) for crystal form transformation of the active pharmaceutical ingredient within that time window.

[0029] In this embodiment, it should be specifically noted that the time window should be a continuous and fixed-duration monitoring period. Crystal form transformation is a dynamic process that requires continuous time series data to capture and accumulate risks. Only continuous anomalies within the time window can reflect the true transformation trend. The time window should be selected according to the applicable scenario. For the R&D stage, the time window can be set to 30 minutes to capture risk fluctuations in the process at high frequency. For the mass production stage, the time window should match the batch cycle, such as 2 hours, to fit the production rhythm and facilitate process traceability.

[0030] S3: Crystal form transformation risk assessment: Preset a comprehensive risk threshold for crystal form transformation, determine the risk level of crystal form transformation based on the comprehensive risk index of crystal form transformation, and identify key risk influencing factors; Furthermore, the comprehensive risk threshold for crystal form transformation includes the crystal form transformation safety threshold (CRI). s Crystal form transformation early warning threshold CRI w and the CRI (Crystal Form Transformation Risk Threshold) o Based on the comprehensive risk index of crystal form transformation and the preset comprehensive risk threshold of crystal form transformation, the risk level of crystal form transformation is determined, including the first risk level, the second risk level, the third risk level and the fourth risk level.

[0031] In this embodiment, it should be specifically noted that the steps for judging the risk level of crystal form transformation are as follows: B1: When CRI > CRI s the crystal form transformation is at the first risk level, indicating a green risk-free state. All parameters are within the stable range and there is no driving force for transformation; B2: When CRI s < CRI < CRI w the crystal form transformation is at the second risk level, indicating a yellow warning state. One-dimensional parameter is close to the threshold; B3: When CRI w < CRI < CRI o the crystal form transformation is at the third risk level, indicating an orange high-risk state. The probability of crystal form transformation is greater than 30%; B4: When CRI o < CRI, the crystal form transformation is at the fourth risk level, indicating a red out-of-control state and the crystal form has been transformed.

[0032] Furthermore, the identification of risk key influencing factors needs to be based on the comprehensive risk index of crystal form transformation. By decomposing the sub-index and attributing the SHAP value, the key parameters leading to risks are identified, the contribution degrees of the crystal form characteristic risk index, crystal form environment risk index and regulation process risk index are calculated, sorted, and the sub-parameters with high contribution degrees of the risk index are extracted to obtain the influence intensity of each sub-parameter on the risk index. The sub-parameter with the highest influence intensity is used as the risk key influencing factor.

[0033] In this embodiment, it should be specifically noted that the number of risk key influencing factors can be multiple. When the detected risk level of crystal form transformation is relatively high, the SHAP algorithm needs to be used to analyze the risk indexes of the three dimensions, calculate the influence degree of each parameter on the risk index, select 3 - 5 parameters with relatively high influence degrees as the risk key influencing factors. Substituting the risk key influencing factors into the historical risk case library for verification means that if more than 80% of the same type of risk batches have abnormal values of this parameter, it is determined as a true risk key influencing factor.

[0034] S4: Generation of adjustment control instructions: Based on the preset regulation rules, the intelligent control terminal generates production equipment regulation instructions according to the risk level and risk key influencing factors; Furthermore, the steps for generating production equipment regulation instructions are as follows: C1: Locate the risk key influencing factors and match the specific regulation authorities based on the regulation rules; C2: Dynamically calculate the regulation amount of the risk key influencing factors and perform instruction verification; C3: Generates executable PLC code, including control permissions and control values, and adds audit trail tags.

[0035] In this embodiment, it should be specifically explained that the preset control rule refers to matching the corresponding equipment control authority according to the crystal form conversion risk level and key risk influencing factors. The preset control rule needs to be continuously optimized in combination with risk cases. When a certain control instruction successfully reduces the risk index, the weight of the rule should be strengthened. If the risk index does not improve after control, the rule correction process is triggered.

[0036] S5: Issuance and execution of control instructions: The intelligent control terminal transmits control instructions for the production equipment to the production equipment, executes the control instructions, and provides real-time feedback on the control effect data. Furthermore, the control effect data includes equipment status data, process parameter data, and crystal form characteristic data. Equipment status data includes equipment operating parameters, energy consumption, and execution sequence. Process parameter data includes thermodynamic parameters, kinetic parameters, and environmental parameters. Crystal form characteristic data includes structural parameters, morphological parameters, and stability parameters.

[0037] In this embodiment, it is necessary to specifically explain that the equipment operating parameters include, but are not limited to, cooling pump speed and valve opening; energy consumption includes, but is not limited to, compressor power and steam flow rate; execution time includes, but is not limited to, instruction reception time, start execution time and completion time; thermodynamic parameters include, but are not limited to, crystallization vessel temperature and pressure; kinetic parameters include, but are not limited to, supersaturation and crystal nucleus growth rate; environmental parameters include, but are not limited to, workshop humidity and cleanliness; structural parameters include, but are not limited to, XRD characteristic peak intensity ratio and Raman peak shift; morphological parameters include, but are not limited to, crystal aspect ratio and surface roughness; and stability parameters include, but are not limited to, crystal purity and metastable crystal form ratio.

[0038] It should be specifically noted that monitoring equipment status data is used to verify whether the instructions are correctly received and executed by the equipment; monitoring process parameter data is used to monitor whether the process status changes in the expected direction after the instructions are executed; and monitoring crystal form characteristic data is used to directly verify whether the crystal form transforms into a more stable state.

[0039] S6: Feedback-optimized closed-loop control: The data processing terminal updates the crystal form transformation risk assessment model and control rules based on the control effect data.

[0040] Furthermore, updating the crystal form transformation risk assessment model and control rules requires the data processing terminal to evaluate the control effectiveness index, parameter response speed, and crystal form stability index based on the real-time feedback control effect data. Root cause analysis is then performed based on the evaluation results, and the crystal form transformation risk assessment model is updated and the control rules are optimized based on the evaluation results.

[0041] In this embodiment, it should be specifically noted that the calculation of the regulation effectiveness index requires calculating the CRI change rate after regulation. If it is negative, it means that the risk has been effectively reduced. The larger the absolute value, the more obvious the risk reduction effect. The parameter response speed refers to the time from the issuance of the regulation command to the target value, reflecting the equipment execution efficiency. The crystal form stability index refers to the fluctuation coefficient of the crystal form characteristic parameter after 1 hour of regulation. The smaller the coefficient, the more stable the crystal form of the active pharmaceutical ingredient.

[0042] It should be specifically noted that the update of the crystal form transformation risk assessment model includes the adjustment of model parameter weights, the adaptive adjustment of the three-dimensional risk index calculation formula, and the dynamic calibration of safety and risk thresholds; the optimization of the regulation rules includes the addition and elimination of the rule base, the optimization of the regulation amount calculation model, and the differentiated adaptation of rules for multiple scenarios.

[0043] Secondly: The accompanying drawings of the embodiments disclosed in this invention only involve the structures involved in the embodiments disclosed in this invention. Other structures can refer to the general design. In the absence of conflict, the same embodiment and different embodiments of this invention can be combined with each other. In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An intelligent analysis method for the crystal form transformation risk of a pharmaceutical active ingredient, characterized in that, This includes an online detection terminal, a data processing terminal, and an intelligent control terminal. All terminals are connected via a real-time communication protocol, and the steps are as follows: S1: Data Acquisition and Preprocessing: Crystal form transformation related parameters are acquired in real time through an online detection terminal, and then transmitted to the data processing terminal after data preprocessing; S2: Comprehensive Data Analysis and Evaluation: The data processing terminal constructs a crystal form transformation risk assessment model and calculates the comprehensive crystal form transformation risk index based on the crystal form transformation-related parameters. S3: Crystal form transformation risk assessment: Preset a comprehensive risk threshold for crystal form transformation, determine the risk level of crystal form transformation based on the comprehensive risk index of crystal form transformation, and identify key risk influencing factors; S4: Adjustment of control command generation: Based on preset control rules, the intelligent control terminal generates control commands for production equipment according to the risk level and key risk influencing factors; S5: Issuance and execution of control instructions: The intelligent control terminal transmits control instructions for the production equipment to the production equipment, executes the control instructions, and provides real-time feedback on the control effect data. S6: Feedback-optimized closed-loop control: The data processing terminal updates the crystal form transformation risk assessment model and control rules based on the control effect data.

2. The intelligent analysis method for the crystal form transformation risk of a pharmaceutical active ingredient according to claim 1, characterized in that: The crystal form transformation correlation parameters include crystal form characteristic parameters, crystal form environment parameters, and control process parameters. The crystal form characteristic parameters include spectral quantification parameters, diffraction quantification parameters, and morphology quantification parameters. The crystal form environment parameters include temperature and humidity quantification parameters and gas parameters. The control process parameters include material process parameters and time and energy parameters.

3. The intelligent analysis method for the crystal form transformation risk of a pharmaceutical active ingredient according to claim 1, characterized in that: The steps for constructing the crystal form transformation risk assessment model are as follows: A1: Parameter screening and weight allocation: Screen out key parameters related to crystal form transformation from the parameters related to crystal form transformation, and assign weights to each parameter; A2: Parameter Standardization: Using the standardization formula: , Mapping the key parameters associated with crystal form transformation to the [0,1] interval, X i S represents the key parameter associated with the i-th crystal form transformation. i X represents the standardized key parameter associated with the i-th crystal form transformation. i安全阈值 X represents the safety threshold of the key parameter associated with the i-th crystal form transformation. i危险阈值 This represents the danger threshold of the key parameter associated with the i-th crystal form transformation; A3: Calculation of risk indicators by dimension: Based on the standardized parameters, the crystal form characteristic risk index FRI, the crystal form environmental risk index ERI, and the regulation process risk index PRI are calculated. A4: Comprehensive Risk Index Coupling: Based on the crystal form characteristic risk index, crystal form environmental risk index, and control process risk index, the following formula is used: , The comprehensive risk index (CRI) for crystal form transformation was calculated. A5: Model Weight Correction: Verify and correct the model weights.

4. The intelligent analysis method for the crystal form transformation risk of a pharmaceutical raw material according to claim 1, characterized in that: To obtain the comprehensive risk index of crystal form transformation, a time window needs to be set, the crystal form transformation-related parameters within the time window need to be obtained, the crystal form transformation-related parameters need to be imported into the crystal form transformation risk assessment model, and the comprehensive risk index of crystal form transformation of the active pharmaceutical ingredient (API) within the time window needs to be calculated.

5. The intelligent analysis method for the crystal form transformation risk of a pharmaceutical raw material according to claim 1, characterized in that: The comprehensive risk threshold for crystal form transformation includes the crystal form transformation safety threshold (CRI). s Crystal form transformation early warning threshold CRI w and the CRI (Crystal Form Transformation Risk Threshold) o Based on the comprehensive risk index of crystal form transformation and the preset comprehensive risk threshold of crystal form transformation, the risk level of crystal form transformation is determined, including the first risk level, the second risk level, the third risk level and the fourth risk level.

6. The intelligent analysis method for the crystal form transformation risk of a pharmaceutical active pharmaceutical ingredient according to claim 1, characterized in that: The identification of key risk influencing factors requires the use of a comprehensive risk index for crystal form transformation. By decomposing sub-indices and attributing SHAP values, key parameters that dominate the risk are identified. The contribution of the crystal form characteristic risk index, crystal form environmental risk index, and regulation process risk index are calculated, ranked, and sub-parameters of risk indices with high contribution are extracted. The influence intensity of each sub-parameter on the risk index is obtained, and the sub-parameter with the highest influence intensity is taken as the key risk influencing factor.

7. The intelligent analysis method for the crystal form transformation risk of a pharmaceutical raw material according to claim 1, characterized in that: The steps for generating the production equipment control commands are as follows: C1: Identify key risk factors and match specific regulatory authority with regulatory rules; C2: Dynamically calculate the adjustment amount of key risk influencing factors and perform instruction verification; C3: Generates executable PLC code, including control permissions and control values, and adds audit trail tags.

8. The intelligent analysis method for the crystal form transformation risk of a pharmaceutical raw material according to claim 1, characterized in that: The control effect data includes equipment status data, process parameter data, and crystal form characteristic data. Equipment status data includes equipment operating parameters, energy consumption, and execution sequence. Process parameter data includes thermodynamic parameters, kinetic parameters, and environmental parameters. Crystal form characteristic data includes structural parameters, morphological parameters, and stability parameters.

9. The intelligent analysis method for the crystal form transformation risk of a pharmaceutical raw material according to claim 1, characterized in that: The updating of the crystal form transformation risk assessment model and control rules requires the data processing terminal to evaluate the control effectiveness index, parameter response speed and crystal form stability index based on the real-time feedback control effect data, perform root cause analysis based on the evaluation results, and update the crystal form transformation risk assessment model and optimize the control rules based on the evaluation results.

Images