Multi-dimensional evaluation method and system for production quality of traditional Chinese medicine, and computer equipment

Abstract

The invention relates to the technical field of traditional Chinese medicine production, and particularly discloses a multi-dimensional evaluation method and system for traditional Chinese medicine production quality and computer equipment. Performing data processing on the production data to obtain to-be-evaluated data; performing risk identification based on the to-be-evaluated data by using a pre-trained model in a machine learning model library, and outputting a risk identification result; and performing score prediction based on the to-be-evaluated data and the risk identification result by using a pre-trained model in a machine learning model library, and outputting a multi-dimensional score prediction value. By introducing machine learning, multi-dimensional comprehensive analysis of production quality in the traditional Chinese medicine production process can be realized, the risk management capability and the quality control accuracy are improved, and the reliability of the traditional Chinese medicine production process is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of traditional Chinese medicine production, and in particular to a multi-dimensional evaluation method, system and computer equipment for traditional Chinese medicine production quality. Background Art

[0002] Quality management methods in traditional Chinese medicine production primarily rely on experience and the monitoring of single-dimensional data. Inspectors evaluate the quality of TCM production by manually sampling and testing product samples based on industry standards, analyzing and evaluating certain physical and chemical indicators (such as moisture, ingredient content, active ingredients, and microbial content). However, standard-based manual testing suffers from inefficiency, difficulty in real-time evaluation, and high subjectivity, which can lead to bias. Summary of the Invention

[0003] Based on this, it is necessary to provide a multi-dimensional evaluation method, system and computer equipment for the production quality of traditional Chinese medicine to address the above problems.

[0004] A multidimensional evaluation method for the production quality of traditional Chinese medicine comprises the following steps: obtaining production data during the production process of the traditional Chinese medicine; processing the production data to obtain data to be evaluated; performing risk identification based on the data to be evaluated using a pre-trained model in a machine learning model library, and outputting a risk identification result; performing score prediction based on the data to be evaluated and the risk identification result using a pre-trained model in the machine learning model library, and outputting a multidimensional score prediction value.

[0005] In one embodiment, the processing of the production data to obtain the data to be evaluated includes cleaning, standardizing, and feature extracting the collected production data to obtain the data to be evaluated.

[0006] In one embodiment, the machine learning model library includes a principal component analysis model, an isolation forest model and a risk assessment model, and the pre-trained model in the machine learning model library is used to identify the risks of the data to be evaluated, and outputting the risk identification results includes using the principal component analysis model to identify the variation pattern of the data to be evaluated; using the isolation forest model to detect and identify abnormal data points in the data to be evaluated; using the risk assessment model to identify high-risk batches and key quality factors in the data to be evaluated; when the data to be evaluated has at least two abnormal factors among the variation pattern, the abnormal data point, the high-risk batch and the key quality factor, outputting the risk identification result of the risk point.

[0007] In one embodiment, when the data to be evaluated contains at least two abnormal factors among the variation pattern, the abnormal data point, the high-risk batch and the key quality factor, after outputting the risk identification result of the risk point, the method further includes outputting warning information based on the abnormal factors existing in the data to be evaluated.

[0008] In one embodiment, the pre-trained model in the machine learning model library is used to score and predict the data to be evaluated and the risk identification result, and the multi-dimensional score prediction value is output, which includes using a deep neural network model to output the score prediction value of the data to be evaluated based on each dimension according to the nonlinear relationship between each dimension and the production data; using a random forest model to evaluate the importance of each feature in the data to be evaluated, and mapping the importance score of each feature in the data to be evaluated to the weight corresponding to each feature; adjusting the weight corresponding to each feature in the data to be evaluated according to the risk identification result; and determining the multi-dimensional score prediction value according to the score prediction value of each dimension and the weight corresponding to each feature.

[0009] A multidimensional evaluation system for the production quality of traditional Chinese medicine comprises a data acquisition module for acquiring production data during the production process of traditional Chinese medicine; a data processing module connected to the data acquisition module for processing the production data to acquire data to be evaluated; a risk identification module connected to the data processing module for performing risk identification based on the data to be evaluated using a pre-trained model in a machine learning model library, and outputting a risk identification result; and a multidimensional score prediction module respectively connected to the data processing module and the risk identification module for performing score prediction based on the data to be evaluated and the risk identification result using a pre-trained model in the machine learning model library, and outputting a multidimensional score prediction value.

[0010] In one embodiment, the multidimensional evaluation system for the production quality of traditional Chinese medicine further includes a feedback module, which is respectively connected to the risk identification module and the multidimensional score prediction module, and is used to display the risk identification results and / or multidimensional score prediction values.

[0011] A computer device includes a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the steps of the multidimensional evaluation method for the production quality of traditional Chinese medicine described in any one of the above embodiments are implemented.

[0012] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the multidimensional evaluation method for the production quality of traditional Chinese medicine described in any one of the above embodiments.

[0013] A computer program product includes a computer program, which, when executed by a processor, implements the steps of the multidimensional evaluation method for the production quality of traditional Chinese medicine described in any one of the above embodiments.

[0014] The multidimensional evaluation method for TCM production quality processes production data from the TCM production process to obtain data to be evaluated. It then uses pre-trained models from a machine learning model library to identify risks in the data to be evaluated, outputting risk identification results. It then uses pre-trained models from the machine learning model library to predict scores for the data to be evaluated and the risk identification results, outputting multidimensional predicted scores. By incorporating machine learning, it is possible to achieve a comprehensive, multidimensional analysis of production quality during TCM production, improving risk management capabilities, the accuracy of quality control, and the reliability of the TCM production process. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the implementation methods of this specification or the technical solutions in the prior art, the following is a brief introduction to the drawings required for use in the implementation methods or the description of the prior art. Obviously, the drawings described below are only some implementation methods recorded in this specification. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0016] Figure 1 This is a schematic diagram of a method flow chart for a multidimensional evaluation method for the production quality of traditional Chinese medicine in one embodiment of the present application;

[0017] Figure 2 This is a flow chart of a method for risk identification based on data to be evaluated in one embodiment of the present application;

[0018] Figure 3 This is a flow chart of a method for performing score prediction based on the data to be evaluated and the risk identification results in one embodiment of the present application;

[0019] Figure 4 This is a schematic diagram of the structure of a multi-dimensional evaluation system for the production quality of traditional Chinese medicine in one embodiment of the present application;

[0020] Figure 5 This is a schematic diagram of the structure of a device for implementing a multi-dimensional evaluation method for the production quality of traditional Chinese medicine in one embodiment of the present application;

[0021] Figure 6 This is a diagram of the internal structure of a computer device in one of the embodiments of the present application. DETAILED DESCRIPTION

[0022] To facilitate understanding of the present invention, the present invention will be described more fully below with reference to the accompanying drawings. Preferred embodiments of the present invention are shown in the accompanying drawings. However, the present invention may be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided to provide a more thorough and comprehensive understanding of the present disclosure.

[0023] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this invention pertains. The terms used in this specification of the present invention are for the purpose of describing specific embodiments only and are not intended to limit the present invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0024] The production process of Traditional Chinese Medicine (TCM) is complex, with multiple intertwined factors potentially affecting product quality, including raw material quality, specific production process parameters, and environmental conditions. These factors are highly nonlinear and multidimensional, creating uncertainty in TCM product quality. Traditional quality control methods are unable to cope with this complexity.

[0025] Figure 1 This is a method flow chart of a multidimensional evaluation method for the production quality of traditional Chinese medicine in one embodiment of the present application. In one embodiment, the multidimensional evaluation method for the production quality of traditional Chinese medicine may include the following steps S100 to S400.

[0026] Step S100: Acquire production data during the production process of traditional Chinese medicine.

[0027] Collect all kinds of production data in the production process of traditional Chinese medicine to achieve all-round data collection. Among them, production data may include but are not limited to equipment operating parameters (such as temperature, humidity, pressure, pH value, chemical composition), process status, finished product quality indicators (such as disintegration time, ingredient content), etc. Specifically, various sensors can be set to collect parameters such as temperature, humidity, and pressure in real time; by building a unified data interface standard, ensure that equipment of different brands and models can be seamlessly connected, and use the data interface to access various production equipment or operating systems to obtain equipment information in real time; use various collection devices to obtain all kinds of production data that can be collected in real time. By obtaining all kinds of data covering the entire production process in real time, the comprehensiveness and accuracy of the data can be ensured.

[0028] Step S200: Process the production data to obtain data to be evaluated.

[0029] Before evaluating production data, the production data can be processed to facilitate subsequent analysis. In this embodiment, the production data that has completed data processing is defined as data to be evaluated. The data to be evaluated can be used as high-quality data input for further analysis and evaluation.

[0030] Step S300: Use the pre-trained model in the machine learning model library to perform risk identification based on the data to be evaluated, and output the risk identification results.

[0031] In this embodiment, a machine learning model library can be pre-built, and the machine learning model library can include various pre-trained models. According to the characteristics of the data to be evaluated (such as data volume, feature dimension, time series characteristics, etc.) and the specific problems to be solved (such as anomaly detection, classification prediction, trend analysis, etc.), the most suitable model is selected from the machine learning model library. The data to be evaluated is input into the selected machine learning model, and the risk identification task is performed and the risk identification results are output by calling the API or directly loading the local model file. By using a fully trained machine learning model, potential risks can be identified more accurately, and the false alarm rate and missed alarm rate can be reduced, thereby ensuring product quality. The machine learning model has the ability to quickly process large amounts of complex data. Therefore, the use of a pre-trained machine learning model can complete the risk assessment in the shortest time and provide timely warnings of possible problems.

[0032] Step S400: Use the pre-trained model in the machine learning model library to perform score prediction based on the data to be evaluated and the risk identification results, and output a multi-dimensional score prediction value.

[0033] The data to be evaluated generated by step S200 and the risk identification results output by step S300 are integrated into a new feature set. These features are preprocessed as necessary, such as standardization, normalization, feature selection or creation of new composite features to ensure the data quality input into the scoring prediction model. According to the specific needs of the scoring prediction, a suitable model is selected from the machine learning model library. The preprocessed feature data is passed into the selected machine learning model to obtain a multidimensional scoring prediction value. In this embodiment, multidimensional scoring can refer to the evaluation of the production quality of traditional Chinese medicine from multiple different dimensions based on the data to be evaluated. The multidimensional scoring prediction value can integrate scores of multiple dimensions including but not limited to human operation, materials, preparation methods, environment, process parameters, etc.

[0034] The judgment of production quality in the existing traditional Chinese medicine production mainly relies on the experience of practitioners and lacks objective data support, resulting in a highly subjective quality evaluation. At the same time, the existing evaluation methods mainly rely on manual sampling to detect certain indicators (such as active ingredients and microbial content), which makes it difficult to capture the continuous changes and comprehensive impacts of the production process in real time, difficult to handle complex nonlinear data, and unable to dynamically reflect the complex relationship between multiple production variables and multidimensional quality characteristics of the product. In other words, the existing evaluation methods can only achieve single-dimensional monitoring. In addition, the test results are usually obtained after the production is completed. The existing quality evaluation methods are mostly offline tests, which usually require sampling and testing before obtaining results. The test cycle is long and the data integration capability is limited. This lag cannot meet the needs of real-time quality control in the production process, increasing the risk of quality problems in traditional Chinese medicine production.

[0035] The multidimensional evaluation method for the production quality of traditional Chinese medicine provided in this application introduces multi-source data fusion technology, and dynamically integrates and comprehensively analyzes various data sources (such as process parameters, environmental conditions, personnel operation records, material properties, manufacturing process, etc.) in the production process of traditional Chinese medicine. Based on machine learning and statistical analysis methods, real-time quality evaluation means are established through real-time data streams such as sensors to achieve real-time monitoring and evaluation of the production process of traditional Chinese medicine. Advanced nonlinear modeling methods (such as deep learning, etc.) are used to capture the complex relationship between multidimensional data and quality indicators in the production process. By combining risk identification results and multidimensional features, combining historical data and real-time data to perform score prediction, the accuracy and relevance of the score are improved, and it is closer to the actual situation. Efficient calculation process and rapid response mechanism ensure the timeliness of score prediction. Operators can adjust the production process in time based on risk identification results and multidimensional score prediction values to prevent the occurrence of quality problems.

[0036] By integrating multi-source data and combining multi-dimensional factors such as process parameters, environmental conditions, personnel operation records, material properties, and manufacturing processes in the traditional Chinese medicine production process, this approach overcomes the limitations of existing assessment methods that focus on a single dimension, providing a holistic quality evaluation perspective and achieving comprehensive quality evaluation, which can more accurately reflect the overall quality characteristics of traditional Chinese medicine products. Pre-trained models from the machine learning model library are used to enable real-time dynamic monitoring of the production process, providing real-time feedback on quality changes, quickly identifying quality deviations, and providing timely feedback to adjust process parameters, thus avoiding the lag of traditional offline testing and significantly improving production efficiency and quality stability. Using pre-trained models from the machine learning model library to assess production quality and potential risk points in real time can provide forward-looking quality management capabilities, reduce potential quality risks, support intelligent optimization of the production process, reduce resource waste, and provide scientific data support for production process optimization and early prevention of quality issues.

[0037] In one embodiment, processing the production data to obtain the data to be evaluated may include the following steps: cleaning, standardizing, and feature extracting the collected production data to obtain the data to be evaluated.

[0038] High-quality data input is provided by cleaning, standardizing, and feature extracting collected data. Specifically, data cleaners can be used to remove errors and missing values from production data by deleting, replacing, or marking them. Data cleaners can also be used to merge time windows of time series data to ensure data integrity. Standardization tools can be used to make production data comparable through normalization or normal distribution transformation. Feature extraction tools can be used to extract key features from raw time series data to provide strong support for subsequent analysis.

[0039] Preprocessing collected production data through cleaning, standardization, and feature extraction can significantly reduce noise and improve data integrity and accuracy. Feature extraction effectively extracts the most representative information from massive amounts of raw data, providing strong support for subsequent analysis. High-quality features not only improve the model's predictive accuracy but also reveal patterns and regularities hidden in the data. Ensuring the quality of data input to machine learning models provides a solid foundation for multidimensional evaluation methods of Traditional Chinese Medicine production quality.

[0040] In one embodiment, the machine learning model library can include models pre-trained based on historical data, such as principal component analysis (PCA), isolation forest models, and risk assessment models. Using the models in the machine learning model library to perform time series data analysis on the data to be assessed can identify potential risks across multiple dimensions of the production process.

[0041] Figure 2 This is a flow chart of a method for risk identification based on the data to be evaluated in one of the embodiments of the present application, in which the risk of the data to be evaluated is identified using a pre-trained model in a machine learning model library, and outputting the risk identification results may include the following steps S310 to S330.

[0042] Step S310: using the principal component analysis model to identify the variation pattern of the data to be evaluated.

[0043] Principal Component Analysis (PCA) is a statistical method used to reduce dimensionality and identify key patterns of variation in data. PCA can transform high-dimensional data into a low-dimensional representation while preserving the original data's variance as much as possible. This is crucial for understanding data structure, discovering underlying patterns, and reducing noise.

[0044] Specifically, the covariance matrix between all features is calculated, which reflects the strength of the linear relationship between the features. The covariance matrix is eigendecomposed to obtain a set of eigenvalues and their corresponding eigenvectors. The eigenvalue represents the proportion of data variance explained by each principal component; the eigenvector defines the direction of the new coordinate system. According to the cumulative contribution rate (that is, the proportion of the total variance explained by the first few principal components), an appropriate number of principal components are selected to construct a new low-dimensional representation. The data to be evaluated is projected into a new coordinate system composed of the selected principal components to obtain a simplified data representation. Use visualization tools such as scatter plots and heat maps to display the principal component scores to help intuitively understand the main variation patterns of the data.

[0045] In a preferred embodiment, as the traditional Chinese medicine production process progresses, new production data is continuously added, and the PCA model can be retrained regularly to ensure that it adapts to the latest production conditions.

[0046] Step S320: Detect and identify abnormal data points in the data to be evaluated using the isolation forest model.

[0047] Isolation Forest is an unsupervised anomaly detection method for continuous data. Starting from an outlier, it divides the data using specified rules and makes judgments based on the number of divisions. It constructs multiple isolation trees by randomly selecting features and split points, isolating outliers earlier in the tree and thus identifying anomalous data points.

[0048] Specifically, the number of isolation trees to be constructed is specified, the number of samples to be used for each isolation tree is determined, and the proportion of outliers is estimated. A threshold is set based on actual needs, below which data points are considered outliers. This threshold can be optimized through methods such as cross-validation. The trained isolation forest model is applied to the data to be evaluated, outputting an anomaly score for each data point. A lower score indicates a higher likelihood of an outlier.

[0049] In a preferred embodiment, the isolation forest model can also be retrained regularly to establish a feedback loop, and the model parameters and threshold settings can be continuously optimized according to the results of actual production and quality control to ensure that it adapts to the latest production conditions.

[0050] Step S330: using the risk assessment model to identify high-risk batches and critical quality factors in the data to be assessed.

[0051] Using risk assessment models to identify high-risk batches and critical quality factors (CQFs) within the data under evaluation is a crucial step in ensuring product quality and safety during the production of traditional Chinese medicine. This can be achieved by integrating various machine learning techniques and statistical methods. A risk score is calculated for each batch using a trained classification model, with a higher score indicating a greater likelihood of quality issues. For CQFs, the impact of each feature on risk can be determined based on its importance score (e.g., feature importance using random forests).

[0052] Step S340: When the data to be evaluated has at least two abnormal factors among variation patterns, abnormal data points, high-risk batches and key quality factors, a risk identification result of the risk points is output.

[0053] Risk assessment is performed on the data to be evaluated simultaneously through multiple anomaly detection methods to obtain abnormal data together. When two or more abnormal factors appear in multiple anomaly detection methods, it can be judged as abnormal and it can be considered that there are potential risk points in the data to be evaluated, and the risk identification results of the risk points are output.

[0054] In one embodiment, when the data to be evaluated contains at least two abnormal factors among variation patterns, abnormal data points, high-risk batches and key quality factors, after outputting the risk identification result of the risk point, the method may further include the following steps: outputting warning information based on the abnormal factors existing in the data to be evaluated.

[0055] When risk points are detected in the data to be evaluated, warning messages can be generated based on the abnormal factors in the data to ensure the safety of the production process and product quality. For example, if abnormal temperature fluctuations are detected in certain batches of products, a corresponding warning message can be automatically generated: "Abnormal temperature fluctuations have been detected on a certain production line during a certain period of time. There may be quality issues. Please check immediately."

[0056] In the embodiments of the present application, based on machine learning and statistical analysis methods, a real-time risk assessment model is established through real-time data streams such as sensors, which enables real-time monitoring and assessment of risks in the production process of traditional Chinese medicine, thereby improving risk management capabilities and the accuracy of quality control. By achieving real-time dynamic monitoring of the production process of traditional Chinese medicine, the lag of traditional offline detection is avoided, and potential risk points in the production process of traditional Chinese medicine can be quickly identified. In addition, timely feedback is provided so that operators can adjust process parameters to avoid quality problems and promote continuous quality management and improvement. The detection and adjustment of quality problems are completed before the product is shipped to avoid unnecessary waste and quality risks.

[0057] Figure 3This is a flow chart of a method for performing score prediction based on the data to be evaluated and the risk identification results in one embodiment of the present application. In one embodiment, obtaining the appointment status information of the target inspection time period may include the following steps S410 to S440.

[0058] Step S410: Use a deep neural network model to output a predicted score value of the data to be evaluated based on each dimension according to the nonlinear relationship between each dimension and the production data.

[0059] Deep Neural Networks (DNNs) are machine learning models based on artificial neural networks. They feature multiple layers of nonlinear transformation units and are designed to learn complex representations of data, including complex nonlinear mapping relationships. DNNs can accurately capture the complex nonlinear relationships between multidimensional data and quality indicators during the production process, enabling quality assessment of multidimensional data from the traditional Chinese medicine production process.

[0060] Specifically, design a DNN architecture suitable for processing structured or time-series data, such as a multilayer perceptron (MLP), convolutional neural network (CNN), or recurrent neural network (RNN / LSTM). Define the input layer based on the number and type of features to ensure that all features are correctly passed to the model. The model is trained based on historical data and establishes a prediction and evaluation framework by learning the complex nonlinear relationships between the data. Define the output layer to match the needs of rating prediction. For example, for multi-dimensional rating prediction, multiple output nodes can be set, each corresponding to the rating of a dimension. Select an appropriate loss function and optimization algorithm to adjust the model parameters. Use historical data to train the DNN model to ensure that the model can learn the patterns in the data. Apply the trained DNN model to predict the data to be evaluated, and output the predicted rating values based on each dimension.

[0061] Step S420: Use the random forest model to evaluate the importance of each feature in the data to be evaluated, and map the importance score of each feature in the data to be evaluated to a weight corresponding to each feature.

[0062] Random forest refers to a classifier that uses multiple trees to train and predict samples. It is an ensemble learning method primarily used for classification and regression tasks. It is a model composed of multiple decision trees that improves the accuracy and stability of the model by integrating the prediction results of these decision trees. The random forest model can accurately assess the importance of each feature, thereby identifying the key factors that have the greatest impact on product quality. The random forest model itself has good interpretability. Combined with feature importance analysis and visualization tools, it can clearly reveal which factors have a significant impact on the score, increasing confidence. The weights derived based on feature importance can help management better understand weak links in the production process and assist in the development of more effective quality control strategies.

[0063] Specifically, a random forest model is constructed, and a classifier or regressor is selected based on the nature of the problem. A trained random forest model directly provides importance scores for each feature. These scores reflect the contribution of each feature across all decision trees, typically based on Gini impurity or information gain. Feature importance scores can be normalized to the range [0, 1], allowing them to be directly used for weight assignment. Feature importance scores are mapped to weights, ensuring that the sum of all weights equals 1, as required by the probability distribution. By analyzing the weights of each feature, the key quality factors that have the greatest impact on the score can be identified.

[0064] Step S430: Adjust the weights corresponding to the various features in the data to be evaluated according to the risk identification results.

[0065] Based on real-time data feedback (i.e., risk identification results), weight allocations are dynamically adjusted to ensure the scoring model is adapted to current production conditions. Operators can dynamically adjust weights based on the company's accumulated experience and product focus.

[0066] Step S440: Determine a multi-dimensional score prediction value based on the score prediction value of each dimension and the weight corresponding to each feature.

[0067] In the score prediction values for each dimension output in step S410, different quality indicators in different dimensions have different prediction values. In this step, the multi-dimensional score prediction value can be determined based on the weights corresponding to each quality indicator in each dimension. By aggregating the scores of various indicators, an overall quality score can be formed.

[0068] This application provides a multidimensional evaluation method for the production quality of traditional Chinese medicine. By integrating advanced real-time data acquisition technology and machine learning algorithms, a comprehensive dynamic evaluation of multidimensional quality factors is achieved. Through multi-source data fusion technology, multidimensional factors such as process parameters, environmental conditions, personnel operation records, material properties, and manufacturing process in the traditional Chinese medicine production process are combined. The value and weight of each quality indicator are determined through machine learning, providing a scientific multidimensional (personnel operation, materials, manufacturing method, environment, process parameters) comprehensive score to achieve comprehensive quality evaluation.

[0069] In one embodiment, comprehensive monitoring and feedback can also be performed on real-time data in the production process to ensure that product quality meets expectations. Specifically, a monitoring dashboard can be used to display real-time quality scores and the status of important indicators. The quality score can be a multi-dimensional score of the process data by the multi-dimensional score prediction module 400. The important indicators can refer to key quality factors that have a greater impact on the score, such as the temperature on the extraction tank, the temperature of the distilled condensate, the pressure in the extraction tank, the flow rate of the extracted liquid, the cumulative amount of solvent, the cumulative amount of aromatic water, etc. An alarm is used to automatically alert the identified risks or abnormalities to prompt the operator. By designing a feedback mechanism unit, the feedback mechanism unit is used to transmit the monitoring results back to the production line. The monitoring results can serve as a theoretical basis for the operator to guide the real-time adjustment and optimization of the production process. Among them, the monitoring results can include risk identification results and multi-dimensional score prediction values.

[0070] In a preferred embodiment, the machine learning model library may also include a dynamic prediction model. The dynamic prediction model can combine real-time data (i.e., real-time data to be evaluated) and historical data to achieve accurate prediction of the quality change trend of traditional Chinese medicine. The quality change trend of traditional Chinese medicine is used to provide scientific guidance for production process optimization and early prevention of quality problems. Furthermore, based on the quality change trend of traditional Chinese medicine output by the dynamic prediction model, optimization suggestions for quality changes can also be provided, providing forward-looking quality management capabilities, reducing potential quality risks, supporting intelligent optimization of the production process, and reducing resource waste.

[0071] It should be understood that although the steps in the flowcharts of the accompanying drawings are shown in sequence as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and these steps can be executed in other orders. Moreover, at least some of the steps in the flowcharts of the accompanying drawings may include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times. The order of execution of these steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least a portion of the steps or stages in other steps.

[0072] Based on the description of the embodiment of the multidimensional evaluation method for the production quality of traditional Chinese medicine mentioned above, the present disclosure also provides a multidimensional evaluation system for the production quality of traditional Chinese medicine. The system may include a system (including a distributed system), software (application), modules, components, servers, clients, etc. using the method described in the embodiments of this specification and combined with necessary implementation hardware devices. Based on the same innovative concept, the system in one or more embodiments provided by the embodiments of the present disclosure is as described in the following embodiments. Since the implementation scheme and method for solving the problem of the system are similar, the implementation of the specific system of the embodiments of this specification can refer to the implementation of the aforementioned method, and the repetitions will not be repeated. As used below, the term "unit" or "module" can implement a combination of software and / or hardware for predetermined functions. Although the system described in the following embodiments is preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceived.

[0073] Figure 4 This is a structural diagram of a multidimensional evaluation system for the production quality of traditional Chinese medicine in one of the embodiments of the present application. In one embodiment, the multidimensional evaluation system for the production quality of traditional Chinese medicine may include a data acquisition module 100, a data processing module 200, a risk identification module 300 and a multidimensional score prediction module 400.

[0074] The data acquisition module 100 can be used to acquire production data during the traditional Chinese medicine production process. The data acquisition module 100 can collect various types of production data during the traditional Chinese medicine production process, achieving comprehensive data collection. Production data may include, but is not limited to, equipment operating parameters (such as temperature and pressure), process status, and finished product quality indicators (such as disintegration time and ingredient content).

[0075] In one feasible implementation, the data acquisition module 100 may include sensors, data interfaces, and related hardware. Various sensors can be used to collect parameters such as temperature, humidity, and pressure in real time. The data interfaces connect to various production devices or operating systems to obtain real-time device information. Furthermore, various acquisition devices can be configured to acquire other types of collectible production data in real time.

[0076] The data processing module 200 can be connected to the data acquisition module 100 and can be used to process the production data to obtain data to be evaluated. Before evaluating the production data, the data processing module 200 can first process the production data to facilitate subsequent analysis.

[0077] In a feasible implementation, the data processing module 200 may include a data cleaner, a standardization tool, and a feature extraction tool. The data processing module 200 may clean, standardize, and extract features from the collected data to provide high-quality data input.

[0078] The risk identification module 300 can be connected to the data processing module 200. The risk identification module 300 can be used to use pre-trained models in the machine learning model library to identify risks based on the data to be evaluated and output risk identification results. In this embodiment, a machine learning model library can be pre-built and can include various pre-trained models. The risk identification module 300 can select the most suitable model from the machine learning model library based on the characteristics of the data to be evaluated and the specific problem to be solved. The data to be evaluated is input into the selected machine learning model, and the risk identification task is executed and the risk identification results are output by calling the API or directly loading the local model file.

[0079] The multidimensional rating prediction module 400 can be connected to the data processing module 200 and the risk identification module 30 respectively. The multidimensional rating prediction module 400 can be used to use the pre-trained model in the machine learning model library to perform rating prediction based on the data to be evaluated and the risk identification results, and output the multidimensional rating prediction value.

[0080] In a feasible implementation, the multidimensional scoring prediction module 400 can integrate the data to be evaluated output by the data processing module 200 and the risk identification results output by the risk identification module 300 into a new feature set. These features are subjected to necessary preprocessing, such as standardization, normalization, feature selection or creation of new composite features, to ensure the quality of data input into the scoring prediction model. The multidimensional scoring prediction module 400 can select a suitable model from the machine learning model library according to the specific needs of the scoring prediction. The preprocessed feature data is passed into the selected machine learning model to obtain a multidimensional scoring prediction value. In this embodiment, multidimensional scoring can refer to the evaluation of the production quality of traditional Chinese medicine from multiple different dimensions based on the data to be evaluated. The multidimensional scoring prediction value can integrate scores of multiple dimensions including but not limited to human operation, materials, preparation methods, environment, process parameters, etc.

[0081] The multidimensional evaluation system for TCM production quality provided in this application utilizes a modular design, encompassing multiple aspects of control, including data collection, processing, risk identification, score calculation, and monitoring feedback, ensuring efficient, real-time management and quality optimization of the production process. By integrating advanced real-time data collection technology with machine learning algorithms, a comprehensive, dynamic evaluation of multidimensional quality factors is achieved.

[0082] In one embodiment, the multi-dimensional evaluation system for the production quality of traditional Chinese medicine may further include a feedback module.

[0083] The feedback module can be connected to the risk identification module 300 and the multi-dimensional score prediction module 400, respectively. The feedback module can be used to display the risk identification results and / or the multi-dimensional score prediction values. Comprehensive monitoring and feedback of real-time data in the production process ensure that product quality meets expectations.

[0084] In a feasible embodiment, the feedback module may include a monitoring dashboard, an alarm and a feedback mechanism unit. Among them, the monitoring dashboard can display real-time quality scores and important indicator status. The quality score can be a multi-dimensional score of the process data by the multi-dimensional score prediction module 400, and the important indicators can refer to key quality factors that have a greater impact on the score, such as the temperature on the extraction tank, the temperature of the distilled condensate, the pressure in the extraction tank, the flow rate of the extracted liquid, the cumulative amount of solvent, the cumulative amount of aromatic water, etc. The alarm can automatically alert the identified risks or abnormalities to prompt the operator. The feedback mechanism unit can transmit the monitoring results back to the production line, and the monitoring results can serve as a theoretical basis for the operator to guide the real-time adjustment and optimization of the production process. Among them, the monitoring results can include risk identification results and multi-dimensional score prediction values.

[0085] It should be understood that the various embodiments of the above-mentioned methods, systems, etc. in this specification are described in a progressive manner. The same / similar parts between the various embodiments can be referred to in detail. Each embodiment focuses on the differences from other embodiments. For related parts, please refer to the descriptions of other method embodiments.

[0086] Figure 5 This is a schematic diagram of the device structure for implementing a multi-dimensional evaluation method for the production quality of traditional Chinese medicine in one embodiment of this application. Figure 5 The multidimensional evaluation system S00 for the production quality of traditional Chinese medicines may include a processing component S20, which further includes one or more processors and a memory resource represented by a memory S22 for storing instructions executable by the processing component S20, such as an application program. The application program stored in the memory S22 may include one or more modules, each corresponding to a set of instructions. Furthermore, the processing component S20 is configured to execute the instructions to perform the multidimensional evaluation method for the production quality of traditional Chinese medicines described above.

[0087] The multi-dimensional evaluation system for Chinese medicine production quality S00 may further include: a power supply component S24 configured to perform power management of the multi-dimensional evaluation system for Chinese medicine production quality S00, a wired or wireless network interface S26 configured to connect the multi-dimensional evaluation system for Chinese medicine production quality S00 to a network, and an input / output (I / O) interface S28. The multi-dimensional evaluation system for Chinese medicine production quality S00 may operate based on an operating system stored in the memory S22, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, or the like.

[0088] In an exemplary embodiment, a computer-readable storage medium including instructions is further provided, such as a memory S22 including instructions. The instructions can be executed by a processor of the multidimensional evaluation system S00 for the production quality of traditional Chinese medicine to perform the above method. The storage medium can be a computer-readable storage medium, such as a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, or the like.

[0089] In an exemplary embodiment, a computer program product is further provided. The computer program product includes instructions, and the instructions can be executed by a processor of the multi-dimensional evaluation system S00 for the production quality of traditional Chinese medicine to complete the above method.

[0090] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 6 As shown, Figure 6 This is a diagram of the internal structure of a computer device in one embodiment of the present application. The computer device includes a processor, a memory, and a network interface connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The database of the computer device is used to store user- and task-related data used in the above-mentioned multidimensional evaluation method for the production quality of traditional Chinese medicine. The network interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, a multidimensional evaluation method for the production quality of traditional Chinese medicine can be implemented.

[0091] Those skilled in the art will understand that Figure 6 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0092] Those skilled in the art will appreciate that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, database or other media used in the embodiments provided in this application may include at least one of non-volatile and volatile memory. Non-volatile memory may include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory may include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in the various embodiments provided herein may include at least one of a relational database and a non-relational database. Non-relational databases may include, but are not limited to, distributed databases based on blockchains. The processor involved in the various embodiments provided herein may be, but are not limited to, a general-purpose processor, a central processing unit, a graphics processing unit, a digital signal processor, a programmable logic unit, a data processing logic unit based on quantum computing, and the like.

[0093] The various embodiments in this specification are described in a progressive manner. Similar portions between the various embodiments can be referenced to each other. Each embodiment focuses on the differences between the other embodiments. In particular, the hardware + program embodiments are described briefly because they are generally similar to the method embodiments. For relevant portions, refer to the description of the method embodiments.

[0094] It should be noted that the aforementioned systems, electronic devices, servers, etc., according to the description of the method embodiments, may also include other implementation methods. For specific implementation methods, reference can be made to the description of the relevant method embodiments. Furthermore, new embodiments formed by combining features of various method, system, device, and server embodiments remain within the scope of implementation of this disclosure and are not detailed here.

[0095] Throughout this specification, references to terms such as "some embodiments," "other embodiments," and "desired embodiments" indicate that a particular feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. Although these terms are used interchangeably throughout this specification, they do not necessarily refer to the same embodiment or example.

[0096] The technical features of the above-mentioned embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the technical features in the above-mentioned embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0097] The above-described embodiments merely represent several implementations of the present invention. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the patent. It should be noted that a person skilled in the art would be able to make various modifications and improvements without departing from the spirit of the present invention, all of which fall within the scope of protection of the present invention. Therefore, the scope of protection of the patent for this invention shall be determined by the appended claims.

Claims

1. A multidimensional evaluation method for the production quality of traditional Chinese medicine, characterized in that: include: Obtain production data during the production process of traditional Chinese medicine; Processing the production data to obtain data to be evaluated; Use a pre-trained model in the machine learning model library to perform risk identification based on the data to be evaluated and output a risk identification result; A pre-trained model in the machine learning model library is used to perform score prediction based on the data to be evaluated and the risk identification results, and a multi-dimensional score prediction value is output.

2. The multidimensional evaluation method for the production quality of traditional Chinese medicine according to claim 1, characterized in that: Processing the production data to obtain data to be evaluated includes: The collected production data is cleaned, standardized and feature extracted to obtain data to be evaluated.

3. The multidimensional evaluation method for the production quality of traditional Chinese medicine according to claim 1 or 2, characterized in that: The machine learning model library includes a principal component analysis model, an isolation forest model, and a risk assessment model. The pre-trained model in the machine learning model library is used to identify the risk of the data to be assessed, and the output risk identification result includes: Identifying the variation pattern of the data to be evaluated using the principal component analysis model; Detecting and identifying abnormal data points in the data to be evaluated using the isolation forest model; Using the risk assessment model to identify high-risk batches and critical quality factors in the data to be assessed; When the data to be evaluated has at least two abnormal factors among the variation pattern, the abnormal data point, the high-risk batch and the key quality factor, a risk identification result of the risk point is output.

4. The multidimensional evaluation method for the production quality of traditional Chinese medicine according to claim 3, characterized in that: When the data to be evaluated has at least two abnormal factors among the variation pattern, the abnormal data point, the high-risk batch, and the key quality factor, after outputting a risk identification result of the risk point, the method further includes: Output warning information based on the abnormal factors existing in the data to be evaluated.

5. The multidimensional evaluation method for the production quality of traditional Chinese medicine according to claim 1, characterized in that: The use of a pre-trained model in a machine learning model library to perform score prediction on the data to be evaluated and the risk identification result, and outputting a multi-dimensional score prediction value includes: Using a deep neural network model to output a predicted score value of the data to be evaluated based on each dimension according to the nonlinear relationship between each dimension and the production data; Use a random forest model to evaluate the importance of each feature in the data to be evaluated, and map the importance score of each feature in the data to be evaluated to a weight corresponding to each feature; Adjust the weight corresponding to each feature in the data to be evaluated according to the risk identification result; The multi-dimensional score prediction value is determined based on the score prediction value of each dimension and the weight corresponding to each feature.

6. A multidimensional evaluation system for the production quality of traditional Chinese medicine, characterized by: include: Data acquisition module, used to obtain production data during the production process of traditional Chinese medicine; A data processing module, connected to the data acquisition module, for processing the production data to obtain data to be evaluated; A risk identification module, connected to the data processing module, is used to use a pre-trained model in a machine learning model library to perform risk identification based on the data to be evaluated and output a risk identification result; A multidimensional score prediction module is connected to the data processing module and the risk identification module respectively, and is used to use a pre-trained model in the machine learning model library to perform score prediction based on the data to be evaluated and the risk identification results, and output a multidimensional score prediction value.

7. The multidimensional evaluation system for the production quality of traditional Chinese medicine according to claim 6, characterized in that: The multidimensional evaluation system for the production quality of traditional Chinese medicine also includes: The feedback module is connected to the risk identification module and the multi-dimensional score prediction module respectively, and is used to display the risk identification result and / or the multi-dimensional score prediction value.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the multidimensional evaluation method for the production quality of traditional Chinese medicine according to any one of claims 1 to 5 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the multidimensional evaluation method for the production quality of traditional Chinese medicine according to any one of claims 1 to 5 are implemented.

10. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the multidimensional evaluation method for the production quality of traditional Chinese medicine according to any one of claims 1 to 5 are implemented.

Images