A eucommia ulmoides flavone bitter reduction process monitoring and early warning method based on multi-source data statistics

Abstract

The application belongs to the field of intelligent manufacturing equipment application, and discloses a eucommia ulmoides flavone bitter-reducing process monitoring and early warning method based on multi-source data statistics; original raw material quality data, flavone bitter-reducing process parameter characteristics and bitter-reducing eucommia ulmoides chemical component parameter characteristics are collected, high-dimensional feature processing is performed on the original raw material quality data, the flavone bitter-reducing process parameter characteristics and the bitter-reducing eucommia ulmoides chemical component parameter characteristics to obtain eucommia ulmoides flavone bitter-reducing process monitoring characteristics, a bitter-reducing process monitoring and early warning model is constructed by using the monitoring characteristics and the eucommia ulmoides flavone bitter-reducing process monitoring and early warning method with corresponding label processing, and real-time eucommia ulmoides flavone bitter-reducing process monitoring and early warning methods are obtained by using the model processing. The application realizes comprehensive statistical analysis and early warning of multi-source data, and uses an improved machine learning model to output real-time eucommia ulmoides flavone bitter-reducing process monitoring and early warning methods. The bitter-reducing early warning is changed from traditional experience-driven to data model-driven precise early warning.

Description

Technical Field

[0001] This invention belongs to the field of intelligent manufacturing equipment application technology, and in particular relates to a monitoring and early warning method for the process of reducing bitterness of Eucommia ulmoides flavonoids based on multi-source data statistics. Background Technology

[0002] Eucommia flavonoids are a class of flavonoid compounds extracted from Eucommia ulmoides. They are mainly found in the leaves and male flowers of Eucommia ulmoides, with smaller amounts in the bark and fruit. They are one of the important active ingredients of Eucommia ulmoides.

[0003] Currently, more than 10 flavonoids have been isolated from Eucommia ulmoides, mainly quercetin, kaempferol, and their glycosides (such as rutin, isoquercitrin, hyperoside, and astragaloside). The total flavonoid content in different parts is approximately: male flowers 3%–4%, leaves 1.5%–2.5%, bark 0.4%–1.5%, and seeds 0.06%–1.2%, with male flowers and leaves being the main raw materials for development and utilization. Eucommia ulmoides flavonoids have a wide range of biological activities, mainly including: antioxidant and anti-aging: by scavenging free radicals, chelating metal ions, and enhancing the activity of antioxidant enzymes such as superoxide dismutase (SOD) and glutathione peroxidase (GSH-Px), they reduce oxidative stress damage and delay cell aging. Cardiovascular protection: reducing capillary fragility and permeability, improving vascular elasticity; lowering blood lipids and cholesterol, inhibiting platelet aggregation, and preventing thrombosis, they have an adjunctive therapeutic effect on hypertension, coronary heart disease, and angina pectoris.

[0004] However, existing technologies for reducing bitterness in Eucommia ulmoides flavonoids still have problems such as large fluctuations in process parameters, difficulty in quantifying quality indicators, and delayed abnormal early warning. In recent years, big data analysis technology has become increasingly sophisticated. Further exploration is needed on how to integrate multi-source data collection, statistical modeling, and intelligent early warning to build a process quality control system that can be monitored online and given real-time warnings for Eucommia ulmoides flavonoids.

[0005] Therefore, there are still some challenges in how to collect the quality data of raw materials, process the characteristics of the flavonoid bitterness reduction process parameters and the chemical composition parameters of Eucommia ulmoides for bitterness reduction, and construct a multi-source data bitterness reduction process monitoring and early warning model from the three sources of source, process and result. Specifically, how can the model process the above characteristics to obtain a real-time Eucommia ulmoides flavonoid bitterness reduction process monitoring and early warning method? In other words, how can a comprehensive statistical analysis of source, process and result data be used to provide early warning, shifting the bitterness reduction early warning from traditional experience-driven to data-driven and model-driven precision manufacturing? This would help to steadily improve the quality of Eucommia ulmoides flavonoid products, reduce the defect rate, and achieve intelligent manufacturing of Eucommia ulmoides for bitterness reduction. Summary of the Invention

[0006] To address the aforementioned technical problems, this invention proposes a monitoring and early warning method for the bitterness-reducing process of Eucommia ulmoides flavonoids based on multi-source data statistics.

[0007] In a first aspect of the present invention, a method for monitoring and early warning of the bitterness-reducing process of Eucommia ulmoides flavonoids based on multi-source data statistics is provided, the method comprising:

[0008] S1. Retrieve the original raw material quality data stored in historical batches, obtain the characteristics of the flavonoid bitterness reduction process parameters and the characteristics of the Eucommia ulmoides chemical composition parameters after bitterness reduction, and obtain the corresponding manual experience-based monitoring and early warning method for the Eucommia ulmoides flavonoid bitterness reduction process.

[0009] D2. The original raw material quality data, the flavonoid bitterness reduction process parameter characteristics, and the eucommia chemical component parameters after bitterness reduction are subjected to high-dimensional feature processing to obtain the eucommia flavonoid bitterness reduction process monitoring characteristics.

[0010] D3. Construct a bitterness reduction process monitoring and early warning model using the monitoring characteristics of the Eucommia ulmoides flavonoids bitterness reduction process and the monitoring and early warning method of the Eucommia ulmoides flavonoids bitterness reduction process.

[0011] D4. Based on the aforementioned bitterness reduction process monitoring and early warning model, output a real-time Eucommia ulmoides flavonoid bitterness reduction process monitoring and early warning method to monitor and warn of the bitterness reduction status of newly treated batches of Eucommia ulmoides flavonoids.

[0012] Furthermore, the original raw material quality data was obtained by processing the origin, harvesting time, and storage conditions of Eucommia ulmoides.

[0013] Furthermore, the flavonoid bitterness reduction process parameters include high-temperature fermentation parameters and enzymatic hydrolysis parameters. The high-temperature fermentation parameters include fermentation temperature and fermentation time, and the enzymatic hydrolysis parameters include the type of compound enzyme, enzymatic hydrolysis temperature, and enzymatic hydrolysis time.

[0014] Furthermore, the chemical composition parameters of Eucommia ulmoides after bitterness reduction are obtained using the flavonoid content and bitterness score of Eucommia ulmoides after bitterness reduction.

[0015] Furthermore, the high-dimensional feature processing involves constructing a feature vector space from the original raw material quality data, the flavonoid bitterness reduction process parameters, and the eucommia chemical component parameters after bitterness reduction, to obtain the eucommia flavonoid bitterness reduction process monitoring features.

[0016] Furthermore, the bitterness-reducing process monitoring and early warning model adopts a machine learning model based on the improved bitterness-reducing requirements of Eucommia ulmoides flavonoids.

[0017] Furthermore, the machine learning model based on the improved demand for reducing bitterness from Eucommia ulmoides flavonoids adopts the Fisher criterion classifier based on the improved demand for reducing bitterness from Eucommia ulmoides flavonoids.

[0018] A monitoring and early warning system for the bitterness reduction process of Eucommia ulmoides flavonoids based on multi-source data statistics is also provided. This system implements a method for monitoring and early warning of the bitterness reduction process of Eucommia ulmoides flavonoids based on multi-source data statistics. The system includes a raw material quality data storage module, a flavonoid bitterness reduction process parameter retrieval module, a Eucommia ulmoides chemical component parameter feature acquisition module after bitterness reduction, a high-dimensional feature processing module, a bitterness reduction process monitoring and early warning model construction module, and an Eucommia ulmoides flavonoid bitterness reduction status monitoring module.

[0019] The raw material quality data storage module is used to retrieve the raw material quality data stored in historical batches.

[0020] The flavonoid bitterness reduction process parameter retrieval module is used to acquire the characteristics of the flavonoid bitterness reduction process parameters.

[0021] The module for acquiring the chemical component parameters of Eucommia ulmoides after bitterness reduction is used to acquire the chemical component parameters of Eucommia ulmoides after bitterness reduction.

[0022] The high-dimensional feature processing module is used to perform high-dimensional feature processing on the original raw material quality data, the flavonoid bitterness reduction process parameter characteristics, and the bitterness reduction eucommia chemical component parameters to obtain the eucommia flavonoid bitterness reduction process monitoring characteristics.

[0023] The bitterness reduction process monitoring and early warning model construction module: uses the monitoring characteristics of the bitterness reduction process of Eucommia ulmoides flavonoids and the bitterness reduction process monitoring and early warning method of Eucommia ulmoides flavonoids to construct a bitterness reduction process monitoring and early warning model;

[0024] The Eucommia flavonoid bitterness reduction monitoring module: Based on the bitterness reduction process monitoring and early warning model, it outputs a real-time Eucommia flavonoid bitterness reduction process monitoring and early warning method to monitor and warn of the bitterness reduction status of newly processed batches of Eucommia flavonoids.

[0025] Furthermore, the bitterness-reducing process monitoring and early warning model adopts a machine learning model based on the improved bitterness-reducing requirements of Eucommia ulmoides flavonoids.

[0026] Furthermore, the machine learning model based on the improved demand for reducing bitterness from Eucommia ulmoides flavonoids adopts the Fisher criterion classifier based on the improved demand for reducing bitterness from Eucommia ulmoides flavonoids.

[0027] Therefore, the beneficial effects of this invention are as follows: by collecting raw material quality data, flavonoid bitterness-reducing process parameter characteristics, and bitterness-reducing Eucommia ulmoides chemical component parameter characteristics, high-dimensional feature processing is performed on the raw material quality data, flavonoid bitterness-reducing process parameter characteristics, and bitterness-reducing Eucommia ulmoides chemical component parameter characteristics to obtain Eucommia ulmoides flavonoid bitterness-reducing process monitoring characteristics. A bitterness-reducing process monitoring and early warning model is constructed using the monitoring characteristics and a corresponding tagging method. The model is then used to obtain a real-time Eucommia ulmoides flavonoid bitterness-reducing process monitoring and early warning method. This invention provides early warning through comprehensive statistical analysis of multi-source data and uses an improved machine learning model to output a real-time Eucommia ulmoides flavonoid bitterness-reducing process monitoring and early warning method. This shifts the bitterness-reducing early warning from traditional experience-driven to data model-driven precise early warning.

[0028] Further embodiments and improvements of the present invention will be described in conjunction with the accompanying drawings and specific examples. Attached Figure Description

[0029] Figure 1 This is a flowchart of the method for monitoring and early warning of the bitterness reduction process of Eucommia ulmoides flavonoids based on multi-source data statistics according to the present invention;

[0030] Figure 2 This is a schematic diagram of the monitoring and early warning method system for the bitterness reduction process of Eucommia ulmoides flavonoids based on multi-source data statistics according to the present invention;

[0031] Figure 3 This is a partial classification result diagram of the improved Fisher classifier in an embodiment of the present invention;

[0032] Figure 4 This is a schematic diagram of an electronic device structure for implementing the method of the present invention in an embodiment of the present invention. Detailed Implementation

[0033] The invention will now be further described in conjunction with the accompanying drawings and specific embodiments. The Fisher criterion classifier model used in this invention is an improved model for processing corresponding high-dimensional input features after specific modifications to the Fisher criterion classifier model in the context of this application.

[0034] The Fisher criterion classification model of this invention is used for monitoring and early warning of the bitterness reduction process of Eucommia ulmoides flavonoids. It belongs to the category of intelligent monitoring and early warning of agricultural products and is applied to the intelligent manufacturing equipment industry for reducing bitterness of Eucommia ulmoides flavonoids. Therefore, it belongs to the intelligent manufacturing equipment industry.

[0035] In a first aspect of the present invention, a method for monitoring and early warning of the bitterness-reducing process of Eucommia ulmoides flavonoids based on multi-source data statistics is provided, the method comprising:

[0036] S1. Retrieve the original raw material quality data stored in historical batches, obtain the characteristics of the flavonoid bitterness reduction process parameters and the characteristics of the Eucommia ulmoides chemical composition parameters after bitterness reduction, and obtain the corresponding manual experience-based monitoring and early warning method for the Eucommia ulmoides flavonoid bitterness reduction process.

[0037] D2. The original raw material quality data, the flavonoid bitterness reduction process parameter characteristics, and the eucommia chemical component parameters after bitterness reduction are subjected to high-dimensional feature processing to obtain the eucommia flavonoid bitterness reduction process monitoring characteristics.

[0038] D3. Construct a bitterness reduction process monitoring and early warning model using the monitoring characteristics of the Eucommia ulmoides flavonoids bitterness reduction process and the monitoring and early warning method of the Eucommia ulmoides flavonoids bitterness reduction process.

[0039] D4. Based on the aforementioned bitterness reduction process monitoring and early warning model, output a real-time Eucommia ulmoides flavonoid bitterness reduction process monitoring and early warning method to monitor and warn of the bitterness reduction status of newly treated batches of Eucommia ulmoides flavonoids.

[0040] Eucommia flavonoids are a class of flavonoid compounds extracted from the bark, leaves, and other parts of the Eucommia ulmoides tree. They possess antioxidant, anti-inflammatory, and metabolic-regulating effects and are one of the core components of Eucommia ulmoides' medicinal value. Their active ingredients mainly include quercetin and kaempferol, and they are widely used in traditional Chinese medicine and health products. In this invention, the raw material can be Eucommia ulmoides leaves or bark; unless otherwise specified in this example, it refers to eucommia flavonoids extracted from Eucommia ulmoides leaves.

[0041] Furthermore, the original raw material quality data was obtained by processing information on the origin, harvesting time, and storage conditions of Eucommia ulmoides, and the calculation formula is as follows:

[0042]

[0043] In the formula, R represents the quality data of the original raw material. Set values ​​for the origin of the original raw materials. Optimal raw material origin settings The time of harvesting the original raw materials. The optimal time for harvesting raw materials. This indicates deviation from the optimal harvest time. A value of 1 indicates deviation; otherwise, a value of 0.6 indicates deviation. This data is a rough estimate set by technicians based on the quality of raw materials obtained at different harvest times during Eucommia ulmoides flavonoid extraction. n represents the total storage days. The average storage temperature of Eucommia ulmoides on day i is given. This is the optimal storage temperature for Eucommia ulmoides.

[0044] In this embodiment, the origin of the raw material Eucommia flavonoids from Wangcang, Cili and Lueyang is set as 3, the origin of other regions such as Lingbao is set as 2, Zunyi in Guizhou and the Yunnan-Guizhou Plateau region is set as 1, and other regions are set according to actual needs. This is a data-based setting of the raw material quality data to facilitate data processing in the subsequent model.

[0045] In this embodiment, the quality of the original Eucommia ulmoides raw material before extracting Eucommia ulmoides flavonoids is obtained, so that the bitterness reduction process of Eucommia ulmoides flavonoids under different quality Eucommia ulmoides will show different results and be dynamically corresponded, thereby realizing a high degree of accuracy in monitoring and early warning of the bitterness reduction process of Eucommia ulmoides flavonoids.

[0046] Furthermore, the flavonoid bitterness reduction process parameters include high-temperature fermentation parameters and enzymatic hydrolysis parameters. The high-temperature fermentation parameters include fermentation temperature and fermentation time, and the enzymatic hydrolysis parameters include the type of compound enzyme, enzymatic hydrolysis temperature, and enzymatic hydrolysis time.

[0047] Since fermentation and enzymatic hydrolysis are the most important processes in reducing bitterness in Eucommia ulmoides flavonoids, this application only selects high-temperature fermentation parameters and enzymatic hydrolysis parameters as the characteristic parameters of the flavonoid bitterness reduction process in the model construction.

[0048] In this embodiment, the bitterness is reduced by microbial fermentation of Eucommia ulmoides leaves or bark at a suitable temperature, which metabolizes or transforms some bitter substances. For example, after high-temperature pretreatment, Eucommia ulmoides leaves are fermented at a constant temperature of 20–60℃ for 3–20 minutes until the leaves turn black, effectively removing the bitter taste while minimizing the loss of active ingredients such as flavonoids and chlorogenic acid. The high-temperature fermentation parameters can be 35℃ and 15 minutes. These process parameters are automated settings for the processing parameters of the processing plant when reducing the bitterness of Eucommia ulmoides flavonoids. Similarly, enzymatic hydrolysis parameters can be added during the extraction of Eucommia ulmoides flavonoids to perform enzymatic hydrolysis, destroy the cell wall structure, increase the flavonoid dissolution rate, and degrade some bitter precursor substances. For example, adding cellulase or pectinase to the extraction of Eucommia ulmoides leaves and performing enzymatic hydrolysis at different temperatures for different times can significantly increase the flavonoid yield and improve the taste. In this embodiment, different types of compound enzymes are set as corresponding values ​​to facilitate data processing and calculation in the subsequent model. According to the needs of model calculation, in this embodiment, cellulase is set to 1 and pectinase is set to 5. If cellulase is set, its enzymatic hydrolysis temperature and enzymatic hydrolysis time are 52℃ and 1 hour respectively.

[0049] Therefore, a characteristic of a flavonoid bitterness reduction process parameter in this embodiment can be represented as (35,15,1,52,1), which corresponds to the corresponding high-temperature fermentation parameters: fermentation temperature, fermentation time, type of compound enzyme, enzymatic hydrolysis temperature, and enzymatic hydrolysis time.

[0050] Furthermore, the chemical composition parameters of Eucommia ulmoides after bitterness reduction were obtained using the flavonoid content and bitterness score of Eucommia ulmoides after bitterness reduction:

[0051]

[0052] In the formula, To characterize the chemical composition parameters of Eucommia ulmoides after bitterness reduction, in this embodiment, these parameters are calculated using the number of Eucommia ulmoides flavonoid samples after the bitterness reduction process, where m represents the number of Eucommia ulmoides flavonoid samples after the bitterness reduction process. The flavonoid content of the j-th Eucommia ulmoides flavonoid sample was obtained by ultraviolet spectrophotometry and high-performance liquid chromatography. Let denot j be the bitterness score of the eucommia flavonoid sample. The bitterness score is obtained through a subjective bitterness scoring method, such as the conventional quinine sulfate solution method, which will not be elaborated here. All the above data have been dimensionless to facilitate subsequent model calculations.

[0053] Furthermore, the high-dimensional feature processing involves constructing a feature vector space from the original raw material quality data, the flavonoid bitterness reduction process parameters, and the eucommia chemical component parameters after bitterness reduction, to obtain the eucommia flavonoid bitterness reduction process monitoring features.

[0054] The feature vector space of the monitoring features of the Eucommia ulmoides flavonoid bitterness reduction process in this embodiment is represented as follows:

[0055]

[0056] To facilitate subsequent model processing, the eigenvector values ​​in the blank areas of the eigenvector space are set to 0. This is a conventional way to supplement eigenvectors. Filling spatial eigenvectors with 0 essentially unifies eigenvectors of different dimensions into the same high-dimensional space, thereby satisfying the requirements for dimensional consistency in matrix operations, model inputs, etc.

[0057] Furthermore, the bitterness-reducing process monitoring and early warning model adopts a machine learning model based on the improved bitterness-reducing requirements of Eucommia ulmoides flavonoids.

[0058] Furthermore, the machine learning model based on the improved demand for bitterness reduction using Eucommia ulmoides flavonoids employs a Fisher criterion classifier based on the improved demand for bitterness reduction using Eucommia ulmoides flavonoids, and its calculation formula is as follows:

[0059]

[0060] In the formula, O represents the monitoring characteristic of the eucommia flavonoid bitterness reduction process. The method for monitoring and early warning of the bitterness reduction process of Eucommia ulmoides flavonoids is to be output. The normal vector perpendicular to the hyperplane is obtained through training using the monitoring features and early warning methods of the Eucommia ulmoides flavonoids-based bitterness-reducing process. The optimal eucommia flavonoid content corresponding to the need for reducing bitterness in eucommia flavonoids. To retain the maximum value of Eucommia flavonoid content under the bitterness-reducing process of Eucommia flavonoids. To retain the minimum flavonoid content of Eucommia ulmoides under the bitterness-reducing process.

[0061] In this embodiment, the optimal eucommia flavonoid content corresponding to the bitterness-reducing effect of eucommia leaves is 1%, and the optimal eucommia flavonoid content corresponding to the bitterness-reducing effect of eucommia bark is 0.85%. This setting is based on data classification and optimization obtained by those skilled in the art through experience and model processing. In this embodiment, a monitoring and early warning method for the bitterness-reducing process using eucommia flavonoids is implemented. Different levels of early warning methods are used to obtain the corresponding classification result values. When the corresponding classification result value is greater than 3.6, it corresponds to a Level 1 warning (alert): the parameter is approaching the control limit or showing slight fluctuations, prompting operators to pay attention. When the corresponding classification result value is less than or equal to 3.6 and greater than or equal to 0, it corresponds to a level 2 warning: the parameter continues to deviate or shows an abnormal trend, triggering process adjustment suggestions (such as adjusting the enzymatic hydrolysis temperature, enzymatic hydrolysis time, etc.). When the corresponding classification result value is less than 0, it corresponds to a level 3 warning: the parameter continues to deviate or shows an abnormal trend, triggering process adjustment suggestions (such as adjusting the enzymatic hydrolysis temperature, enzymatic hydrolysis time, etc.); the parameter seriously exceeds the limit or the quality index is abnormal, automatically triggering shutdown or switching to the backup process scheme to prevent batch scrapping.

[0062] Fisher's criterion classifier is a classic linear discriminant classification method. Its core idea is to find a projection direction that separates samples of different classes as much as possible while clustering samples of the same class as much as possible. Compared with the traditional Fisher's criterion classifier, the classifier in this application has been processed for special application scenarios and has been experimentally verified. It has higher classification results, thereby improving the automation and real-time level of monitoring and early warning of the bitterness reduction process of Eucommia ulmoides flavonoids.

[0063] A monitoring and early warning system for the bitterness reduction process of Eucommia ulmoides flavonoids based on multi-source data statistics is also provided. This system implements a method for monitoring and early warning of the bitterness reduction process of Eucommia ulmoides flavonoids based on multi-source data statistics. The system includes a raw material quality data storage module, a flavonoid bitterness reduction process parameter retrieval module, a Eucommia ulmoides chemical component parameter feature acquisition module after bitterness reduction, a high-dimensional feature processing module, a bitterness reduction process monitoring and early warning model construction module, and an Eucommia ulmoides flavonoid bitterness reduction status monitoring module.

[0064] The raw material quality data storage module is used to retrieve the raw material quality data stored in historical batches.

[0065] The flavonoid bitterness reduction process parameter retrieval module is used to acquire the characteristics of the flavonoid bitterness reduction process parameters.

[0066] The module for acquiring the chemical component parameters of Eucommia ulmoides after bitterness reduction is used to acquire the chemical component parameters of Eucommia ulmoides after bitterness reduction.

[0067] The high-dimensional feature processing module is used to perform high-dimensional feature processing on the original raw material quality data, the flavonoid bitterness reduction process parameter characteristics, and the bitterness reduction eucommia chemical component parameters to obtain the eucommia flavonoid bitterness reduction process monitoring characteristics.

[0068] The bitterness reduction process monitoring and early warning model construction module: uses the monitoring characteristics of the bitterness reduction process of Eucommia ulmoides flavonoids and the bitterness reduction process monitoring and early warning method of Eucommia ulmoides flavonoids to construct a bitterness reduction process monitoring and early warning model;

[0069] The Eucommia flavonoid bitterness reduction monitoring module: Based on the bitterness reduction process monitoring and early warning model, it outputs a real-time Eucommia flavonoid bitterness reduction process monitoring and early warning method to monitor and warn of the bitterness reduction status of newly processed batches of Eucommia flavonoids.

[0070] Furthermore, the bitterness-reducing process monitoring and early warning model adopts a machine learning model based on the improved bitterness-reducing requirements of Eucommia ulmoides flavonoids.

[0071] Furthermore, the machine learning model based on the improved demand for bitterness reduction using Eucommia ulmoides flavonoids employs a Fisher criterion classifier based on the improved demand for bitterness reduction using Eucommia ulmoides flavonoids, and its calculation formula is as follows:

[0072]

[0073] In the formula, O represents the monitoring characteristic of the eucommia flavonoid bitterness reduction process. The method for monitoring and early warning of the bitterness reduction process of Eucommia ulmoides flavonoids is to be output. The normal vector perpendicular to the hyperplane is obtained through training using the monitoring features and early warning methods of the Eucommia ulmoides flavonoids-based bitterness-reducing process. The optimal eucommia flavonoid content corresponding to the need for reducing bitterness in eucommia flavonoids. To retain the maximum value of Eucommia flavonoid content under the bitterness-reducing process of Eucommia flavonoids. To retain the minimum flavonoid content of Eucommia ulmoides under the bitterness-reducing process.

[0074] In this embodiment, the optimal eucommia flavonoid content corresponding to the bitterness-reducing effect of eucommia ulmoides leaves is 1%, and the optimal eucommia flavonoid content corresponding to the bitterness-reducing effect of eucommia ulmoides bark is 0.85%. This setting is based on experience and data obtained through model processing by those skilled in the art for classification and optimization. In this embodiment, as... Figure 3 As shown, the monitoring and early warning method for the bitterness reduction process of Eucommia ulmoides flavonoids is based on the output. Different levels of early warning methods are used to obtain the corresponding classification result values. When the corresponding classification result value is greater than 3.6, it corresponds to a Level 1 warning (alert): the parameter is approaching the control limit or showing slight fluctuations, prompting operators to pay attention. When the corresponding classification result value is less than or equal to 3.6 and greater than or equal to 0, it corresponds to a level 2 warning: the parameter continues to deviate or shows an abnormal trend, triggering process adjustment suggestions (such as adjusting the enzymatic hydrolysis temperature, enzymatic hydrolysis time, etc.). When the corresponding classification result value is less than 0, it corresponds to a level 3 warning: the parameter continues to deviate or shows an abnormal trend, triggering process adjustment suggestions (such as adjusting the enzymatic hydrolysis temperature, enzymatic hydrolysis time, etc.); the parameter seriously exceeds the limit or the quality index is abnormal, automatically triggering shutdown or switching to the backup process scheme to prevent batch scrapping.

[0075] Therefore, the beneficial effects of this invention are achieved by collecting raw material quality data, flavonoid bitterness-reducing process parameters, and bitterness-reducing Eucommia ulmoides chemical component parameters. High-dimensional feature processing is then performed on these raw material quality data, flavonoid bitterness-reducing process parameters, and Eucommia ulmoides chemical component parameters to obtain Eucommia ulmoides flavonoid bitterness-reducing process monitoring features. Using these monitoring features and a corresponding tagging method, a bitterness-reducing process monitoring and early warning model is constructed. This model is then used to obtain a real-time Eucommia ulmoides flavonoid bitterness-reducing process monitoring and early warning method. This invention provides early warning through comprehensive statistical analysis of multi-source data and utilizes an improved machine learning model to output a real-time Eucommia ulmoides flavonoid bitterness-reducing process monitoring and early warning method. This shifts the bitterness-reducing early warning from traditional experience-driven to data model-driven, precise early warning.

[0076] Of course, it is understood that each embodiment of the present invention can achieve one of the effects individually, and the combination of multiple embodiments of the present invention can achieve all the above effects. However, it is not required that each embodiment of the present invention achieve all the above advantages and effects, because each embodiment of the present invention can constitute a separate technical solution and make one or more contributions to the prior art. Each embodiment of the present invention is not affected, and the solution can still be implemented even if one embodiment is deleted.

[0077] For any module structures not specifically defined in this invention, the existing technical descriptions shall prevail. The prior art mentioned in the foregoing background and specific embodiments sections can be considered part of this invention and used to understand the meaning of certain technical features or parameters.

Claims

1. A method for monitoring and early warning of the bitterness-reducing process of Eucommia ulmoides flavonoids based on multi-source data statistics, characterized in that, The method includes: S1. Retrieve the original raw material quality data stored in historical batches, obtain the characteristics of the flavonoid bitterness reduction process parameters and the characteristics of the Eucommia ulmoides chemical composition parameters after bitterness reduction, and obtain the corresponding manual experience-based monitoring and early warning method for the Eucommia ulmoides flavonoid bitterness reduction process. D2. The original raw material quality data, the flavonoid bitterness reduction process parameter characteristics, and the eucommia chemical component parameters after bitterness reduction are subjected to high-dimensional feature processing to obtain the eucommia flavonoid bitterness reduction process monitoring characteristics. D3. Construct a bitterness reduction process monitoring and early warning model using the monitoring characteristics of the Eucommia ulmoides flavonoids bitterness reduction process and the monitoring and early warning method of the Eucommia ulmoides flavonoids bitterness reduction process. D4. Based on the aforementioned bitterness reduction process monitoring and early warning model, output a real-time Eucommia ulmoides flavonoid bitterness reduction process monitoring and early warning method to monitor and warn of the bitterness reduction status of newly treated batches of Eucommia ulmoides flavonoids.

2. The method for monitoring and early warning of the bitterness reduction process of Eucommia ulmoides flavonoids based on multi-source data statistics as described in claim 1, characterized in that: The original raw material quality data was obtained by processing the origin, harvesting time, and storage conditions of Eucommia ulmoides.

3. The method for monitoring and early warning of the bitterness reduction process of Eucommia ulmoides flavonoids based on multi-source data statistics as described in claim 1, characterized in that: The flavonoid bitterness reduction process parameters include high-temperature fermentation parameters and enzymatic hydrolysis parameters. The high-temperature fermentation parameters include fermentation temperature and fermentation time, and the enzymatic hydrolysis parameters include the type of compound enzyme, enzymatic hydrolysis temperature, and enzymatic hydrolysis time.

4. A method for monitoring and early warning of the bitterness reduction process of Eucommia ulmoides flavonoids based on multi-source data statistics as described in claim 2 or 3, characterized in that: The chemical composition parameters of Eucommia ulmoides after bitterness reduction were obtained using the flavonoid content and bitterness score of Eucommia ulmoides after bitterness reduction.

5. The method for monitoring and early warning of the bitterness reduction process of Eucommia ulmoides flavonoids based on multi-source data statistics as described in claim 4, characterized in that: The high-dimensional feature processing involves constructing a feature vector space from the original raw material quality data, the flavonoid bitterness reduction process parameters, and the eucommia chemical component parameters after bitterness reduction, to obtain the monitoring features of the eucommia flavonoid bitterness reduction process.

6. The method for monitoring and early warning of the bitterness reduction process of Eucommia ulmoides flavonoids based on multi-source data statistics as described in claim 5, characterized in that: The bitterness-reducing process monitoring and early warning model adopts a machine learning model based on the improved bitterness-reducing requirements of Eucommia ulmoides flavonoids.

7. The method for monitoring and early warning of the bitterness reduction process of Eucommia ulmoides flavonoids based on multi-source data statistics as described in claim 6, characterized in that: The machine learning model based on the improved demand for reducing bitterness from Eucommia ulmoides flavonoids uses a Fisher criterion classifier based on the improved demand for reducing bitterness from Eucommia ulmoides flavonoids.

8. A monitoring and early warning system for the bitterness reduction process of Eucommia ulmoides flavonoids based on multi-source data statistics. This system implements a method for monitoring and early warning of the bitterness reduction process of Eucommia ulmoides flavonoids based on multi-source data statistics. The system includes a raw material quality data storage module, a flavonoid bitterness reduction process parameter retrieval module, a Eucommia ulmoides chemical component parameter feature acquisition module after bitterness reduction, a high-dimensional feature processing module, a bitterness reduction process monitoring and early warning model construction module, and an Eucommia ulmoides flavonoid bitterness reduction status monitoring module. Its features are: The raw material quality data storage module is used to retrieve the raw material quality data stored in historical batches. The flavonoid bitterness reduction process parameter retrieval module is used to acquire the characteristics of the flavonoid bitterness reduction process parameters. The module for acquiring the chemical component parameters of Eucommia ulmoides after bitterness reduction is used to acquire the chemical component parameters of Eucommia ulmoides after bitterness reduction. The high-dimensional feature processing module is used to perform high-dimensional feature processing on the original raw material quality data, the flavonoid bitterness reduction process parameter characteristics, and the bitterness reduction eucommia chemical component parameters to obtain the eucommia flavonoid bitterness reduction process monitoring characteristics. The bitterness reduction process monitoring and early warning model construction module: uses the monitoring characteristics of the bitterness reduction process of Eucommia ulmoides flavonoids and the bitterness reduction process monitoring and early warning method of Eucommia ulmoides flavonoids to construct a bitterness reduction process monitoring and early warning model; The Eucommia flavonoid bitterness reduction monitoring module: Based on the bitterness reduction process monitoring and early warning model, it outputs a real-time Eucommia flavonoid bitterness reduction process monitoring and early warning method to monitor and warn of the bitterness reduction status of newly processed batches of Eucommia flavonoids.

9. The monitoring and early warning system for the bitterness reduction process of Eucommia ulmoides flavonoids based on multi-source data statistics as described in claim 8, characterized in that: The bitterness-reducing process monitoring and early warning model adopts a machine learning model based on the improved bitterness-reducing requirements of Eucommia ulmoides flavonoids.

10. The monitoring and early warning system for the process of reducing bitterness of Eucommia ulmoides flavonoids based on multi-source data statistics as described in claim 9, characterized in that: The machine learning model based on the improved demand for reducing bitterness from Eucommia ulmoides flavonoids uses a Fisher criterion classifier based on the improved demand for reducing bitterness from Eucommia ulmoides flavonoids.

Images