A fermentation quality detection system and method for Citrus aurantium based on colony characteristic analysis
By constructing a model for colonization of wrinkled microorganisms and tracing the origin of microbial communities, the problems of microenvironmental heterogeneity and microbial community compatibility in the quality detection of bitter orange peel fermentation were solved, enabling accurate detection and grading of bitter orange peel fermentation quality, and improving the stability of the fermentation process and the consistency of product quality.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- JIANGXI UNIVERSITY OF TRADITIONAL CHINESE MEDICINE
- Filing Date
- 2026-02-13
- Publication Date
- 2026-05-26
AI Technical Summary
Existing technologies fail to accurately monitor the microenvironmental heterogeneity of colony distribution and folded structure in the quality detection of trifoliate orange fermentation, leading to misjudgment of fermentation quality and instability of the fermentation process. They also lack comprehensive analysis of the compatibility between local characteristic microorganisms and fermentation functions, and cannot provide accurate quality monitoring and process optimization guidance.
By constructing a wrinkled colony colony colony colony colony colony colony colony colony traceability analysis model and combining the shape characteristics of bitter orange peel and colony distribution data, we can quantify abnormal colony distribution and deviation of microbial community function, establish a fermentation quality risk prediction system, and achieve accurate bitter orange peel fermentation quality detection and grading screening.
Accurately identify fermentation unevenness issues, improve the homogeneity and stability of the fermentation process, optimize raw material selection, reduce the risk of fermentation runaway, reduce resource waste, and ensure product quality consistency and production efficiency.
Smart Images
Figure CN121766848B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of fermented poncirus trifoliata management, and particularly to a quality detection system and method for fermented poncirus trifoliata based on colony feature analysis. Background Art
[0002] As a key raw material in traditional Chinese medicine production and food fermentation industry, the fermentation quality of poncirus trifoliata directly affects the content of active ingredients, bioavailability and production batch stability of products. During the natural fermentation process of poncirus trifoliata, it not only depends on the natural colonization of environmental microorganisms to gradually complete the biotransformation of active ingredients and form specific flavors and active substances, but also the complex fold structure on its surface and the unique flora ecosystem in the origin jointly dominate the fermentation process, from the initial colony adhesion and colonization to the dominant flora competition in the middle stage and then to the metabolite accumulation in the later stage. If the colony distribution characteristics and functional states cannot be accurately monitored, it may lead to quality problems such as uneven fermentation, contamination by miscellaneous bacteria or insufficient transformation of active ingredients, which will not only cause waste of raw materials, but also affect the safety and effectiveness of downstream products. Traditional fermentation quality monitoring mostly relies on chemical analysis of end products or empirical judgment, which not only has lag, but also is difficult to quantitatively characterize the dynamic changes of microorganisms during the fermentation process. By using colony feature analysis technology for quality monitoring, it can, with its precise microbial analysis ability, deeply reveal the microecological environment on the surface of poncirus trifoliata, including the deep folds and concave areas that are difficult to reach by manual evaluation, break through the limitations of traditional quality identification, and significantly improve the fineness and forward-looking of quality monitoring.
[0003] However, when evaluating the fermentation quality of poncirus trifoliata in the prior art, the spatial distribution characteristics of colonies are not associated with the microenvironmental heterogeneity caused by the fold structure of poncirus trifoliata. For example, without considering the fold structure, when the colony distribution is abnormal, it is difficult to accurately judge the root cause of colony imbalance and the risk of fermentation deviation, resulting in misjudgment of fermentation quality. At the same time, the prior art also lacks comprehensive analysis of the adaptability between the characteristic flora in the origin and fermentation functions, resulting in the detection results being difficult to comprehensively reflect the stability risk of the fermentation process and unable to provide accurate guidance for raw material screening and process optimization.
[0004] To solve these problems, the present application designs a quality detection system and method for fermented poncirus trifoliata based on colony feature analysis. Summary of the Invention
[0005] In order to overcome the defects and deficiencies existing in the prior art, the present invention provides a quality detection system and method for fermented poncirus trifoliata based on colony feature analysis. By integrating the shape features of poncirus trifoliata and colony distribution data, a fold colony colonization model is constructed to quantify abnormal distribution situations, and combined with the analysis of functional deviation of the flora in the origin and inherent risk assessment, a fermentation quality risk prediction system is established to achieve precise detection and hierarchical screening of the fermentation quality of poncirus trifoliata based on colony features.
[0006] To achieve the above objectives, the present invention adopts the following technical solution:
[0007] In a first aspect, embodiments of the present invention provide a method for detecting the fermentation quality of Citrus aurantium based on colony characteristic analysis, comprising the following steps:
[0008] S1. Obtain the shape characteristics and origin information of the bitter orange peel in the current fermentation batch, and at the same time obtain the colony distribution data of the bitter orange peel during the fermentation process;
[0009] S2. Combining the shape characteristics data and colony distribution data of bitter orange peel, analyze the distribution characteristics of colonies during the fermentation process of bitter orange peel; and based on the analysis results of the colony distribution characteristics, analyze the abnormal colony distribution at the folds of bitter orange peel.
[0010] S3. Combining the origin information data and colony distribution data of Citrus aurantium, analyze the impact of microbial community differences on the fermentation process of Citrus aurantium, and based on the analysis results of the impact of microbial community differences on the fermentation process of Citrus aurantium, analyze the interference of Citrus aurantium origin on the fermentation process.
[0011] S4. Based on the analysis results of abnormal colony distribution at the folds of the bitter orange peel and the analysis results of the interference of the bitter orange peel production area on the fermentation process, the fermentation quality of the bitter orange peel in the current fermentation batch is evaluated.
[0012] S5. Based on the fermentation quality assessment results of the current fermentation batch of bitter orange peel, the bitter orange peel in the current fermentation batch is screened.
[0013] In one implementation of the present invention, step S2 combines the shape characteristic data and colony distribution data of the bitter orange peel to analyze the colony distribution characteristics during the fermentation process; and based on the analysis results of the colony distribution characteristics, analyzes the abnormal colony distribution at the folds of the bitter orange peel, including the following specific steps:
[0014] S21. Obtain the shape characteristics data of the trifoliate orange peel and the colony distribution data during the fermentation process;
[0015] S22. Classify the bitter orange peel in the current fermentation batch according to the different origins of the bitter orange peel, set up an origin sample analysis unit, and perform batch processing of the above-mentioned data after data fusion using a preset sampling strategy.
[0016] S23. Based on the shape characteristics and colony distribution data of Citrus aurantium, a colony colony colony colony analysis model is constructed to quantify the abnormal colony distribution at the folds of Citrus aurantium in each sample analysis unit from different production areas.
[0017] In one implementation of the present invention, step S23 involves constructing a colony colony analysis model for the wrinkled areas of the bitter orange peel within each sample analysis unit. This process quantifies any abnormalities in the colony distribution at the wrinkled locations of the bitter orange peel. Specifically, this includes the following steps:
[0018] S231. Obtain the shape feature data and colony distribution data of Citrus aurantium. Take the average fold depth and fold surface area ratio of Citrus aurantium in each production area sample analysis unit within the historical monitoring period in the shape feature data, and the colony density ratio of the folded area to the non-folded area of Citrus aurantium in each production area sample analysis unit within the historical monitoring period in the colony distribution data as the regression analysis dataset, and divide the regression analysis dataset into a regression analysis training set and a regression analysis validation set.
[0019] S232. Construct a random forest regression model. Use the average fold depth and fold surface area ratio of Citrus aurantium in each origin sample analysis unit in the regression analysis training set as the input features of the random forest regression model. Use the ratio of the colony density of Citrus aurantium in the folded area to the non-folded area in each origin sample analysis unit in the regression analysis training set as the output target of the random forest regression model. Train the random forest regression model to obtain the initial regression analysis model.
[0020] S233. The initial regression analysis model is validated using the regression analysis validation set, and the initial regression analysis model with an accuracy greater than or equal to the preset first model is used as the wrinkled colony colony analysis model for each origin sample analysis unit.
[0021] S234. Obtain the colony colony analysis model of each sample analysis unit from each production area; use the feature importance analysis method to quantify the contribution of each input feature in the colony colony colony analysis model of each sample analysis unit from each production area to the colony density ratio, take the maximum contribution as the microenvironmental influence of each sample analysis unit from each production area, multiply the microenvironmental influence of each sample analysis unit from each production area by the colony density ratio of the folded area and the non-folded area of the bitter orange peel in each sample analysis unit from each production area within the current monitoring period, and obtain the abnormal colony distribution at the folded position of the bitter orange peel in each sample analysis unit from each production area within the current monitoring period.
[0022] In one implementation of the present invention, step S3 combines the origin information data and colony distribution data of the bitter orange peel to analyze the impact of microbial community differences on the fermentation process of the bitter orange peel. Based on the analysis results of the impact of microbial community differences on the fermentation process of the bitter orange peel, the interference of the origin of the bitter orange peel on the fermentation process is analyzed, including the following specific contents:
[0023] S31. Obtain information on the origin of Citrus aurantium and the colony distribution data during the fermentation process;
[0024] S32. Using the principal coordinate analysis algorithm, based on the functional abundance data of the microbial community in the colony distribution data, extract the functional contour distance matrix of the microbial community of Citrus aurantium in each production area sample analysis unit; based on the functional contour distance matrix of the microbial community, analyze the functional deviation state of the microbial community of Citrus aurantium in each production area sample analysis unit.
[0025] S33. Based on the origin information data and colony distribution data of Citrus aurantium, construct an origin microbial community traceability analysis model to quantify the inherent risk of the microbial community of Citrus aurantium in each origin sample analysis unit.
[0026] S34. The functional deviation and inherent risk of the microbial community of Citrus aurantium in each sample analysis unit are Z-score standardized. The functional deviation and inherent risk of the microbial community of Citrus aurantium in each sample analysis unit after standardization are weighted and summed to obtain the interference of the Citrus aurantium origin on the fermentation process for each sample analysis unit.
[0027] In one implementation of the present invention, step S33 involves constructing a source tracing analysis model for the microbial community of the place of origin, quantifying the inherent risk value of the microbial community within each analysis unit of the sample from the place of origin, specifically including the following steps:
[0028] S331. Obtain the origin information data and colony distribution data of Citrus aurantium. Take the soil type code and harvest accumulated temperature of Citrus aurantium corresponding to each origin sample analysis unit in the historical monitoring period in the origin information data, and the relative abundance of pathogenic bacteria indicator bacteria of Citrus aurantium in each origin sample analysis unit in the historical monitoring period in the colony distribution data as the classification analysis dataset, and divide the classification analysis dataset into classification analysis training set and classification analysis validation set.
[0029] S332. Construct a support vector machine classification model. Use the soil type code and harvest accumulated temperature of the trifoliate orange production area corresponding to each production area sample analysis unit in the classification analysis training set as the input features of the support vector machine classification model. Use whether the relative abundance of pathogenic bacteria indicator bacteria of trifoliate orange in each production area sample analysis unit in the classification analysis training set exceeds the safe abundance threshold as the output label of the support vector machine classification model. Train the support vector machine classification model to obtain the initial classification analysis model for each production area sample analysis unit.
[0030] S333. The initial classification analysis model of each origin sample analysis unit is validated by the classification analysis validation set, and the initial classification analysis model with an accuracy greater than or equal to the preset second model is used as the origin microbial community traceability analysis model for each origin sample analysis unit.
[0031] S334. Obtain the origin-based microbial community traceability analysis model for each origin-based sample analysis unit; use the decision function distance analysis method to quantify the contribution of each input feature in the origin-based microbial community traceability analysis model of each origin-based sample analysis unit to the classification decision boundary; use the contribution of the input feature as soil type coding in the origin-based microbial community traceability analysis model of each origin-based sample analysis unit as the origin-based inherent risk weight of each origin-based sample analysis unit; multiply the origin-based inherent risk weight of each origin-based sample analysis unit by the relative abundance of pathogenic indicator bacteria of Citrus aurantium in each origin-based sample analysis unit within the current monitoring period to obtain the inherent risk of the microbial community of Citrus aurantium in each origin-based sample analysis unit within the current monitoring period.
[0032] In one implementation of the present invention, the fermentation quality of the current batch of bitter orange peel is evaluated based on the analysis results of abnormal colony distribution at the folds of the bitter orange peel and the analysis results of the interference of the bitter orange peel's place of origin on the fermentation process, including the following specific contents:
[0033] S41. Obtain the abnormal colony distribution at the folds of the bitter orange peel in each sample analysis unit of the current monitoring period, and at the same time obtain the interference of the bitter orange peel production area of each sample analysis unit of the current monitoring period on the fermentation process.
[0034] S42. The abnormal colony distribution at the folds of the bitter orange peel in each sample analysis unit of the current production area within the current monitoring period, and the interference of the corresponding bitter orange peel production area on the fermentation process are weighted and summed to obtain the fermentation quality risk of bitter orange peel in each sample analysis unit of the current fermentation batch within the current monitoring period.
[0035] In one implementation of the present invention, the fermentation quality assessment results of the current fermentation batch of bitter orange peel are used to screen the bitter orange peel in the current fermentation batch, including the following specific steps:
[0036] S51. Extract the fermentation quality risk of Citrus aurantium from each production area sample unit in the current fermentation batch during the current monitoring period;
[0037] S52. Preset fermentation quality risk threshold, and compare the fermentation quality risk of each sample analysis unit in the current monitoring period with the fermentation quality risk threshold;
[0038] S53. Discard the bitter orange peels in the corresponding origin sample analysis unit where the fermentation quality risk of bitter orange peel is greater than the fermentation quality risk threshold.
[0039] S54. Continue fermentation treatment for the bitter orange peel samples in the corresponding origin sample analysis unit whose fermentation quality risk is less than or equal to the fermentation quality risk threshold.
[0040] Secondly, embodiments of the present invention also provide a fermentation quality detection system for Citrus aurantium based on colony characteristic analysis, comprising:
[0041] The data acquisition module is used to acquire shape characteristic data and origin information data of the current fermentation batch of bitter orange peel, and at the same time acquire colony distribution data during the fermentation process of bitter orange peel;
[0042] The wrinkled colony detection module is used to analyze the distribution characteristics of colonies during the fermentation process of bitter orange peel by combining the shape feature data and colony distribution data; and to analyze the abnormal colony distribution at the wrinkled position of bitter orange peel based on the analysis results of the colony distribution characteristics.
[0043] The origin interference detection module is used to combine the origin information data and colony distribution data of Citrus aurantium to analyze the impact of microbial community differences on the fermentation process of Citrus aurantium, and to analyze the interference of Citrus aurantium origin on the fermentation process based on the analysis results of the impact of microbial community differences on the fermentation process of Citrus aurantium.
[0044] The fermentation quality analysis module is used to evaluate the fermentation quality of the current batch of bitter orange based on the analysis results of abnormal colony distribution at the folds of the bitter orange peel and the analysis results of the interference of the bitter orange peel origin on the fermentation process.
[0045] The bitter orange peel screening module is used to screen the bitter orange peel in the current fermentation batch based on the fermentation quality assessment results of the current fermentation batch.
[0046] The control module is used to control the operation of the data acquisition module, the wrinkled colony detection module, the origin interference detection module, the fermentation quality analysis module, and the trifoliate orange screening module.
[0047] Compared with the prior art, the present invention has the following advantages and beneficial effects:
[0048] 1. This invention quantifies and analyzes the abnormal colony distribution in the folds of the bitter orange peel, accurately identifies the uneven fermentation problem caused by physical structure, ensures the homogeneity and stability of the fermentation process, and improves the consistency of the final product quality.
[0049] 2. This invention achieves accurate prediction of the fermentation adaptability of raw materials from different origins by comprehensively assessing the deviation state and inherent risks of the microbial community in the place of origin, thereby optimizing the raw material selection criteria and reducing the risk of runaway fermentation process;
[0050] 3. By establishing a multi-parameter fusion quality risk assessment model, this invention can provide early warning of fermentation quality deviations, reduce resource waste caused by the input of inferior raw materials into production, and ensure the controllability of production efficiency and product quality. Attached Figure Description
[0051] Other features, objects, and advantages of the invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:
[0052] Figure 1 This is a schematic diagram of the overall process of the fermentation quality detection method for Citrus aurantium based on colony characteristic analysis of the present invention.
[0053] Figure 2 This is a flowchart of step S2 in the method for detecting the fermentation quality of Citrus aurantium based on colony characteristic analysis of the present invention.
[0054] Figure 3 This is a flowchart of step S3 in the method for detecting the fermentation quality of Citrus aurantium based on colony characteristic analysis of the present invention.
[0055] Figure 4 This is a schematic diagram of the fermentation quality detection system for Citrus aurantium based on colony characteristic analysis according to the present invention. Detailed Implementation
[0056] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the embodiments of the present invention and the specific features in the embodiments are detailed descriptions of the technical solution of the present invention, rather than limitations thereof. In the absence of conflict, the embodiments of the present invention and the technical features in the embodiments can be combined with each other.
[0057] Example 1.
[0058] like Figure 1 As shown, this embodiment provides a method for detecting the fermentation quality of Citrus aurantium based on colony characteristic analysis, specifically including the following steps:
[0059] S1. Obtain the shape characteristics and origin information of the bitter orange peel in the current fermentation batch, and at the same time obtain the colony distribution data of the bitter orange peel during the fermentation process;
[0060] S2. Combining the shape characteristics data and colony distribution data of bitter orange peel, analyze the distribution characteristics of colonies during the fermentation process of bitter orange peel; and based on the analysis results of the colony distribution characteristics, analyze the abnormal colony distribution at the folds of bitter orange peel.
[0061] S3. Combining the origin information data and colony distribution data of Citrus aurantium, analyze the impact of microbial community differences on the fermentation process of Citrus aurantium, and based on the analysis results of the impact of microbial community differences on the fermentation process of Citrus aurantium, analyze the interference of Citrus aurantium origin on the fermentation process.
[0062] S4. Based on the analysis results of abnormal colony distribution at the folds of the bitter orange peel and the analysis results of the interference of the bitter orange peel production area on the fermentation process, the fermentation quality of the bitter orange peel in the current fermentation batch is evaluated.
[0063] S5. Screen the Fructus Aurantii in the current fermentation batch based on the evaluation results of the fermentation quality of Fructus Aurantii in the current fermentation batch.
[0064] In this embodiment, as Figure 2 shown, in step S2, the shape feature data of Fructus Aurantii and the colony distribution data are combined to analyze the distribution characteristic state of colonies during the fermentation process of Fructus Aurantii; and according to the analysis results of the distribution characteristic state of colonies, the abnormal situation of colony distribution at the wrinkled position of Fructus Aurantii is analyzed, including the following specific steps:
[0065] S21. Obtain the shape feature data of Fructus Aurantii and the colony distribution data during the fermentation process;
[0066] S22. Classify the Fructus Aurantii in the current fermentation batch according to different origins of Fructus Aurantii, set an origin sample analysis unit, and batch process the above various types of data after data fusion with a preset sampling strategy; in this embodiment, all samples in the current fermentation batch are classified according to the origin information of Fructus Aurantii, and a sample analysis unit with the origin as the unit is established. This classification method can effectively control the interference of origin factors on the analysis results. In the design of the sampling strategy, this embodiment defaults to using the stratified random sampling method to ensure that the samples in each origin sample analysis unit are fully representative. The data batch processing adopts a parallel computing architecture, and the various types of data after data fusion are uniformly processed through a distributed computing framework, significantly improving the data processing efficiency.
[0067] S23. Based on the shape feature data of Fructus Aurantii and the colony distribution data, construct a wrinkled colony colonization analysis model to quantify the inside of each origin sample analysis unit.
[0068] In this embodiment, in S23, a wrinkled colony colonization analysis model is constructed to quantify the abnormal situation of colony distribution at the wrinkled position of Fructus Aurantii in each origin sample analysis unit, specifically including the following steps:
[0069] S231. Obtain the shape feature data of Fructus Aurantii and the colony distribution data, take the average wrinkling depth and the proportion of wrinkling surface area of Fructus Aurantii in each origin sample analysis unit during the historical monitoring period in the shape feature data, and the colony density ratio between the wrinkled area and the non-wrinkled area of Fructus Aurantii in each origin sample analysis unit during the historical monitoring period in the colony distribution data as the regression analysis data set, and divide the regression analysis data set into a regression analysis training set and a regression analysis validation set;
[0070] S232. Construct a random forest regression model. Use the average fold depth and fold surface area ratio of Citrus aurantium in each origin sample analysis unit in the regression analysis training set as the input features of the random forest regression model. Use the ratio of the colony density of Citrus aurantium in the folded area to the non-folded area in each origin sample analysis unit in the regression analysis training set as the output target of the random forest regression model. Train the random forest regression model to obtain the initial regression analysis model.
[0071] S233. The initial regression analysis model is validated using the regression analysis validation set, and the initial regression analysis model with an accuracy greater than or equal to the preset first model is used as the wrinkled colony colony analysis model for each origin sample analysis unit.
[0072] S234. Obtain the colony colony analysis model of each sample analysis unit from each production area; use the feature importance analysis method to quantify the contribution of each input feature in the colony colony colony analysis model of each sample analysis unit from each production area to the colony density ratio, take the maximum contribution as the microenvironmental influence of each sample analysis unit from each production area, multiply the microenvironmental influence of each sample analysis unit from each production area by the colony density ratio of the folded area and the non-folded area of the bitter orange peel in each sample analysis unit from each production area within the current monitoring period, and obtain the abnormal colony distribution at the folded position of the bitter orange peel in each sample analysis unit from each production area within the current monitoring period.
[0073] It should be noted that this embodiment first establishes a training dataset using historical monitoring data, where the average fold depth and the proportion of folded surface area are used as model input features, and the colony density ratio of folded to non-folded areas is used as the prediction target. The random forest regression model is constructed using the bootstrap sampling method to generate multiple decision trees, and an ensemble learning strategy is used to improve the model's generalization ability. During model training, cross-validation is used to determine the optimal tree depth and node splitting criteria, ensuring that the model can capture complex nonlinear relationships without overfitting. Specifically, in this embodiment, feature importance analysis uses the Gini impurity descent method to accurately quantify the contribution of each input feature to the prediction result. The maximum contribution is defined as the microenvironmental influence parameter, which reflects the critical influence of *Citrus aurantium* folds on colony distribution. Finally, by multiplying the microenvironmental influence by the real-time monitored colony density ratio, abnormal colony distribution is obtained; this accurately captures the complex relationship between *Citrus aurantium* shape characteristics and colony distribution, and identifies key influencing factors through feature importance analysis, providing a scientific basis for subsequent quality assessment and giving the detection results clear biological interpretive significance.
[0074] In this embodiment, step S3 combines the origin information data and colony distribution data of Citrus aurantium to analyze the impact of microbial community differences on the fermentation process of Citrus aurantium. Based on the analysis results of the impact of microbial community differences on the fermentation process of Citrus aurantium, the interference of the origin of Citrus aurantium on the fermentation process is analyzed, including the following specific contents:
[0075] S31. Obtain origin information data and microbial distribution data during fermentation of Citrus aurantium. This embodiment establishes a unified origin information database, systematically collecting key information such as soil type, climate conditions, and cultivation history of each Citrus aurantium origin. Soil type coding adopts the international standard soil classification system to ensure the comparability of origin characteristics. Microbial distribution data is obtained through high-throughput sequencing technology, focusing on the functional abundance characteristics of the microbial community. In the data standardization process, the summation standardization method is used to convert the raw sequencing data into relative abundance data, effectively eliminating the influence of differences in sequencing depth between different samples and making the functional characteristics of the microbial community comparable among different samples. The origin-functional pathway abundance matrix is constructed using a matrix storage method, where rows represent sample units from different origins, columns represent functional pathways in the KEGG database, and matrix element values are the relative abundance values of the corresponding functional pathways. This ensures the systematicness and completeness of the microbial community functional characteristic data, providing standardized data input for subsequent microbial community functional deviation analysis. Simultaneously, the unified database structure allows for longitudinal comparison of test data from different origins and batches, contributing to the establishment of a long-term quality control database.
[0076] S32. Using the principal coordinate analysis algorithm, based on the functional abundance data of the microbial community in the colony distribution data, the functional contour distance matrix of the *Citrus aurantium* microbial community within each origin sample analysis unit is extracted. Based on the functional contour distance matrix, the functional deviation state of the *Citrus aurantium* microbial community within each origin sample analysis unit is analyzed. It should be noted that this embodiment considers both the presence or absence of functional pathways and relative abundance differences, which can comprehensively reflect the differences in the functional composition of the microbial community. The ideal fermentation functional state benchmark is determined by calculation using historical high-quality fermentation batch data, and the mean vector method is used to ensure the representativeness and stability of the benchmark. The construction of the microbial community functional contour distance matrix adopts a symmetric matrix form, which includes both real origin samples and virtual benchmark samples, ensuring spatial consistency in subsequent analysis. The execution of the principal coordinate analysis algorithm projects high-dimensional functional data into a low-dimensional space through eigenvalue decomposition, and the setting of a cumulative variance contribution rate of over 85% ensures the preservation of key information during the dimensionality reduction process. The Euclidean distance is calculated using the geometric distance formula to accurately quantify the degree of deviation of samples from each origin in the functional space. The application of Z-score standardization makes the deviation states of different origins comparable. Based on the above analysis steps, this embodiment establishes a complete calculation chain from microbial community functional characteristics to quality assessment indicators. By quantifying spatial distance, it achieves an intuitive assessment of fermentation quality and provides a reliable quantitative basis for subsequent interference analysis.
[0077] It should be noted that, based on the distance matrix of the microbial community functional profile, the analysis of the deviation state of the microbial community function of *Citrus aurantium* within the analysis unit of samples from each production area specifically includes the following steps:
[0078] S321. Construct an origin-functional pathway abundance matrix based on the functional abundance data of the *Citrus aurantium* community within the analysis unit of each origin sample; and define the benchmark value of the ideal fermentation functional state corresponding to the functional pathway. The benchmark value of the ideal fermentation functional state is set by selecting the mean vector of the origin-functional pathway abundance matrix of historical high-quality fermentation batch samples.
[0079] In this embodiment, the specific process for constructing the origin-functional pathway abundance matrix is as follows: Calculate the sum of the abundance of all functional pathways of the *Citrus aurantium* community in each origin sample analysis unit, and then use the ratio of the abundance of each functional pathway to the sum of the abundance of all functional pathways as the relative abundance of each functional pathway; construct the origin-functional pathway abundance matrix based on the relative abundance of each functional pathway, where the rows in the origin-functional pathway abundance matrix represent *Citrus aurantium* samples, the columns represent functional pathways, and the matrix element values are the relative abundance of each functional pathway;
[0080] S322. Using the Bray-Curtis distance algorithm, the ideal fermentation functional state is used as a virtual sample analysis unit. The dissimilarity of the relative abundance of functional pathways among all sample analysis units from different origins is calculated, as well as the dissimilarity between the relative abundance of functional pathways of all sample analysis units from different origins and the baseline value of the ideal fermentation functional state in the virtual sample analysis unit. A microbial community functional profile distance matrix is constructed among the sample analysis units from different origins. The rows and columns of the microbial community functional profile distance matrix represent all sample analysis units. The matrix element values are the dissimilarity between two sample analysis units in each row and column.
[0081] S323. Based on the distance matrix of the functional contour of the microbial community, execute the principal coordinate analysis algorithm to extract the feature vectors and feature values of the first k principal coordinate axes, and project the functional contour of each sample analysis unit of the production area onto a low-dimensional space based on the principal coordinates.
[0082] In this embodiment, based on the microbial community functional contour distance matrix, the principal coordinate analysis algorithm is executed. The specific process is as follows: the microbial community functional contour distance matrix is subjected to bicentering to convert it into an inner product matrix; the inner product matrix is subjected to eigenvalue decomposition to extract the top k largest eigenvalues and their corresponding eigenvectors. In this embodiment, the value of k must ensure that the cumulative variance contribution rate exceeds 85%; the eigenvectors are multiplied by the square root of the corresponding eigenvalues to obtain the projected coordinates of each sample analysis unit on the principal coordinate axis.
[0083] S324. Extract the projected coordinates of each origin sample analysis unit and the virtual sample analysis unit on the principal coordinate axis, calculate the Euclidean distance between the projected coordinates of each origin sample analysis unit on the principal coordinate axis and the projected coordinates of the virtual sample analysis unit on the principal coordinate axis, and obtain the quantitative value of the functional deviation of the microbial community of Citrus aurantium in each origin sample analysis unit.
[0084] S325. Perform Z-score standardization on the calculated quantitative values of the bacterial community functional deviation of Citrus aurantium in each sample analysis unit from each origin, and output the bacterial community functional deviation of Citrus aurantium in each sample analysis unit from each origin.
[0085] S33. Based on the origin information data and colony distribution data of Citrus aurantium, construct an origin microbial community traceability analysis model to quantify the inherent risk of the microbial community of Citrus aurantium in each origin sample analysis unit.
[0086] S34. The functional deviation and inherent risk of the microbial community of *Citrus aurantium* within each sample analysis unit are Z-score standardized. The standardized microbial functional deviation and inherent risk are then weighted and summed to obtain the interference of the *Citrus aurantium* origin on the fermentation process for each sample analysis unit. This embodiment uses Z-score standardization to eliminate the influence of different indicator dimensions while retaining the relative position information of each indicator in the sample set. The weighted summation calculation employs a dynamic weight allocation mechanism, determining the weight coefficients based on the correlation between each indicator and fermentation quality in historical data, ensuring a high degree of consistency between the evaluation results and the actual quality status. The resulting interference data comprehensively reflects the potential impact of different origins of Citrus aurantium on the standardized fermentation process. This allows for a comprehensive consideration of deviations in microbial community functional characteristics and inherent risk levels, providing important input parameters for subsequent quality assessment. It also enables the quantitative expression of the influence of origin factors. Through the fusion of multi-dimensional indicators, it avoids the limitations of single-indicator assessment, ensuring the comprehensiveness and reliability of the quality assessment results. Furthermore, it provides a clear decision-making basis for the selection of Citrus aurantium raw materials from different origins.
[0087] In this embodiment, in step S33, the construction of the origin-based microbial community tracing analysis model accurately quantifies the inherent risk of the microbial community. This embodiment establishes a classification analysis dataset by collecting origin characteristic data and pathogen indicator bacteria data from historical monitoring periods. The support vector machine classification model employs kernel function techniques to handle nonlinear classification problems, ensuring the model's generalization ability through the principle of minimizing structural risk. During model training, the optimal penalty parameters and kernel function parameters are determined using a grid search method, ensuring the model maintains good classification performance on both the training and validation sets. The decision function distance analysis method used in this embodiment accurately quantifies the contribution of each input feature to the classification decision boundary. The contribution of soil type encoding is defined as the origin-based inherent risk weight, reflecting the decisive influence of different soil types on the microbial community composition. By multiplying the origin-based inherent risk weight by the relative abundance of pathogen indicator bacteria monitored in real time, a quantitative value of the inherent risk of the microbial community is obtained. This accurately identifies the unique microbial community risk characteristics of different origins, uncovers potential risk patterns through machine learning methods, provides a scientific basis for assessing origin disturbance, and makes risk warnings forward-looking and accurate. Specifically, a source tracing analysis model for microbial communities at the place of origin is constructed to quantify the inherent risk value of microbial communities within each sample analysis unit at the place of origin. This includes the following steps:
[0088] S331. Obtain the origin information data and colony distribution data of Citrus aurantium. Take the soil type code and harvest accumulated temperature of Citrus aurantium corresponding to each origin sample analysis unit in the historical monitoring period in the origin information data, and the relative abundance of pathogenic bacteria indicator bacteria of Citrus aurantium in each origin sample analysis unit in the historical monitoring period in the colony distribution data as the classification analysis dataset, and divide the classification analysis dataset into classification analysis training set and classification analysis validation set.
[0089] S332. Construct a support vector machine classification model. Use the soil type code and harvest accumulated temperature of the trifoliate orange production area corresponding to each production area sample analysis unit in the classification analysis training set as the input features of the support vector machine classification model. Use whether the relative abundance of pathogenic bacteria indicator bacteria of trifoliate orange in each production area sample analysis unit in the classification analysis training set exceeds the safe abundance threshold as the output label of the support vector machine classification model. Train the support vector machine classification model to obtain the initial classification analysis model for each production area sample analysis unit.
[0090] S333. The initial classification analysis model of each origin sample analysis unit is validated by the classification analysis validation set, and the initial classification analysis model with an accuracy greater than or equal to the preset second model is used as the origin microbial community traceability analysis model for each origin sample analysis unit.
[0091] S334. Obtain the origin-based microbial community traceability analysis model for each origin-based sample analysis unit; use the decision function distance analysis method to quantify the contribution of each input feature in the origin-based microbial community traceability analysis model of each origin-based sample analysis unit to the classification decision boundary; use the contribution of the input feature as soil type coding in the origin-based microbial community traceability analysis model of each origin-based sample analysis unit as the origin-based inherent risk weight of each origin-based sample analysis unit; multiply the origin-based inherent risk weight of each origin-based sample analysis unit by the relative abundance of pathogenic indicator bacteria of Citrus aurantium in each origin-based sample analysis unit within the current monitoring period to obtain the inherent risk of the microbial community of Citrus aurantium in each origin-based sample analysis unit within the current monitoring period.
[0092] In this embodiment, based on the analysis results of abnormal colony distribution at the folds of the bitter orange peel and the analysis results of the interference of the bitter orange peel's origin on the fermentation process, the fermentation quality of the current batch of bitter orange peel is evaluated, including the following specific contents:
[0093] S41. Obtain the abnormal colony distribution at the folds of the bitter orange peel in each sample analysis unit of the current monitoring period, and at the same time obtain the interference of the bitter orange peel production area of each sample analysis unit of the current monitoring period on the fermentation process.
[0094] S42. The abnormal colony distribution at the folds of the bitter orange peel in each sample analysis unit of the current production area within the current monitoring period, and the interference of the corresponding bitter orange peel production area on the fermentation process are weighted and summed to obtain the fermentation quality risk of bitter orange peel in each sample analysis unit of the current fermentation batch within the current monitoring period.
[0095] It should be noted that the weighted summation calculation in this embodiment uses the weight coefficients determined by the analytic hierarchy process. The relative importance of each indicator is determined by a combination of expert scoring and historical data verification. In this embodiment, the weight of abnormal colony distribution is slightly higher than that of origin interference, which can reflect that the real-time status during fermentation has a stronger immediacy of the impact on quality than inherent characteristics.
[0096] In this embodiment, based on the fermentation quality assessment results of the current fermentation batch of bitter orange peel, the bitter orange peel in the current fermentation batch is screened, including the following specific steps:
[0097] S51. Extract the fermentation quality risk of Citrus aurantium from each production area sample unit in the current fermentation batch during the current monitoring period;
[0098] S52. Preset fermentation quality risk threshold, compare the fermentation quality risk of each production area sample analysis unit with the fermentation quality risk threshold in the current monitoring period; it should be noted that the preset fermentation quality risk threshold in this embodiment is determined by the statistical distribution of historical high-quality batch data, and the safe range is defined by the mean plus or minus two standard deviations to ensure the scientificity and stability of the threshold setting.
[0099] S53. Discard the bitter orange peels in the corresponding origin sample analysis unit where the fermentation quality risk of bitter orange peel is greater than the fermentation quality risk threshold.
[0100] S54. Continue fermentation treatment for the bitter orange peel samples in the corresponding origin sample analysis unit whose fermentation quality risk is less than or equal to the fermentation quality risk threshold.
[0101] In this embodiment, rule engine technology is used to automate the classification and judgment process during the screening decision-making process. Products with quality risks exceeding the threshold are discarded, while products that meet the requirements are allowed to proceed to the subsequent fermentation stage. It is worth noting that this embodiment also establishes a feedback mechanism for screening results, transmitting screening data back to the analysis system in real time for model parameter optimization and dynamic threshold adjustment. The screening execution stage combines machine vision and mechanical sorting equipment to ensure the accuracy and efficiency of the screening operation; it achieves closed-loop quality control management, ensuring the consistency of final product quality through strict screening standards, while automated execution reduces human interference and improves the standardization level of quality control.
[0102] Example 2.
[0103] like Figure 4 As shown, this embodiment provides a fermentation quality detection system for Citrus aurantium based on colony characteristic analysis, including:
[0104] The data acquisition module is used to acquire shape characteristic data and origin information data of the current fermentation batch of bitter orange peel, and at the same time acquire colony distribution data during the fermentation process of bitter orange peel;
[0105] The wrinkled colony detection module is used to analyze the distribution characteristics of colonies during the fermentation process of bitter orange peel by combining the shape feature data and colony distribution data; and to analyze the abnormal colony distribution at the wrinkled position of bitter orange peel based on the analysis results of the colony distribution characteristics.
[0106] The origin interference detection module is used to combine the origin information data and colony distribution data of Citrus aurantium to analyze the impact of microbial community differences on the fermentation process of Citrus aurantium, and to analyze the interference of Citrus aurantium origin on the fermentation process based on the analysis results of the impact of microbial community differences on the fermentation process of Citrus aurantium.
[0107] The fermentation quality analysis module is used to evaluate the fermentation quality of the current batch of bitter orange based on the analysis results of abnormal colony distribution at the folds of the bitter orange peel and the analysis results of the interference of the bitter orange peel origin on the fermentation process.
[0108] The bitter orange peel screening module is used to screen the bitter orange peel in the current fermentation batch based on the fermentation quality assessment results of the current fermentation batch.
[0109] The control module is used to control the operation of the data acquisition module, the wrinkled colony detection module, the origin interference detection module, the fermentation quality analysis module, and the trifoliate orange screening module.
[0110] The parameters and steps for each unit module to achieve the corresponding functions in the above-described quality detection system for fermented bitter orange based on colony feature analysis of the present invention can be referred to the parameters and steps in the embodiments of the method for quality detection of fermented bitter orange based on colony feature analysis described above, and will not be repeated here.
[0111] The various embodiments in this invention are described in a progressive manner. Similar or identical parts between embodiments can be referred to interchangeably. Each embodiment focuses on its differences from other embodiments. In particular, the embodiments for IoT devices and media are relatively simple in description because they are fundamentally similar to the method embodiments; relevant parts can be referred to the descriptions in the method embodiments.
[0112] The systems, media, and methods provided in the embodiments of the present invention are in one-to-one correspondence. Therefore, the systems and media also have similar beneficial technical effects as their corresponding methods. Since the beneficial technical effects of the methods have been described in detail above, the beneficial technical effects of the systems and media will not be repeated here.
[0113] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0114] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.
[0115] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.
[0116] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.
[0117] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.
[0118] Computer-readable media include both permanent and non-permanent, removable and non-removable media that can store information by any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.
[0119] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.
[0120] The above are merely embodiments of the present invention and are not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of the present invention should be included within the scope of the claims of the present invention.
Claims
1. A method for detecting the quality of fermented Fructus Aurantii based on colony characteristic analysis, characterized in that, Includes the following steps: S1. Obtain the shape characteristics and origin information of the bitter orange peel in the current fermentation batch, and at the same time obtain the colony distribution data of the bitter orange peel during the fermentation process; S2. Combining the shape characteristics data and colony distribution data of bitter orange peel, analyze the distribution characteristics of colonies during the fermentation process of bitter orange peel; and based on the analysis results of the colony distribution characteristics, analyze the abnormal colony distribution at the folds of bitter orange peel. The specific steps include the following: S21. Obtain the shape characteristics data of the trifoliate orange peel and the colony distribution data during the fermentation process; S22. Classify the bitter orange peel in the current fermentation batch according to the different origins of the bitter orange peel, set up an origin sample analysis unit, and perform batch processing of various types of data after data fusion using a preset sampling strategy. S23. Based on the shape characteristic data and colony distribution data of Citrus aurantium, construct a colony colony colony analysis model for folds to quantify the abnormal colony distribution at the folds of Citrus aurantium within the analysis units of samples from various production areas; including the following steps: S231. Obtain the shape feature data and colony distribution data of Citrus aurantium. Take the average fold depth and fold surface area ratio of Citrus aurantium in each production area sample analysis unit within the historical monitoring period in the shape feature data, and the colony density ratio of the folded area to the non-folded area of Citrus aurantium in each production area sample analysis unit within the historical monitoring period in the colony distribution data as the regression analysis dataset, and divide the regression analysis dataset into a regression analysis training set and a regression analysis validation set. S232. Construct a random forest regression model. Use the average fold depth and fold surface area ratio of Citrus aurantium in each origin sample analysis unit in the regression analysis training set as the input features of the random forest regression model. Use the ratio of the colony density of Citrus aurantium in the folded area to the non-folded area in each origin sample analysis unit in the regression analysis training set as the output target of the random forest regression model. Train the random forest regression model to obtain the initial regression analysis model. S233. The initial regression analysis model is validated using the regression analysis validation set, and the initial regression analysis model with an accuracy greater than or equal to the preset first model is used as the wrinkled colony colony analysis model for each origin sample analysis unit. S234. Obtain the colony colony analysis model of each sample analysis unit in the production area; use the feature importance analysis method to quantify the contribution of each input feature in the colony colony colony analysis model of each sample analysis unit in the production area to the colony density ratio, take the maximum contribution as the microenvironmental influence of each sample analysis unit in the production area, multiply the microenvironmental influence of each sample analysis unit in the production area by the colony density ratio of the folded area and the non-folded area of the bitter orange peel in each sample analysis unit in the current monitoring period, and obtain the abnormal colony distribution of the folded area of the bitter orange peel in each sample analysis unit in the current monitoring period. S3. Combining the origin information data and colony distribution data of Citrus aurantium, analyze the impact of microbial community differences on the fermentation process of Citrus aurantium, and based on the analysis results of the impact of microbial community differences on the fermentation process of Citrus aurantium, analyze the interference of the origin of Citrus aurantium on the fermentation process; including the following specific contents: S31. Obtain information on the origin of Citrus aurantium and the colony distribution data during the fermentation process; S32. Using the principal coordinate analysis algorithm, based on the functional abundance data of the microbial community in the colony distribution data, extract the functional contour distance matrix of the microbial community of Citrus aurantium in each production area sample analysis unit; based on the functional contour distance matrix of the microbial community, analyze the functional deviation state of the microbial community of Citrus aurantium in each production area sample analysis unit. S33. Based on the origin information data and colony distribution data of Citrus aurantium, construct an origin microbial community traceability analysis model to quantify the inherent risk of the microbial community of Citrus aurantium in each origin sample analysis unit. S34. The functional deviation and inherent risk of the microbial community of Citrus aurantium in each sample analysis unit are Z-score standardized. The functional deviation and inherent risk of the microbial community of Citrus aurantium in each sample analysis unit after standardization are weighted and summed to obtain the interference of the Citrus aurantium origin on the fermentation process in each sample analysis unit. S4. Based on the analysis results of abnormal colony distribution at the folds of the bitter orange peel and the analysis results of the interference of the bitter orange peel production area on the fermentation process, the fermentation quality of the bitter orange peel in the current fermentation batch is evaluated. S5. Based on the fermentation quality assessment results of the current fermentation batch of bitter orange peel, the bitter orange peel in the current fermentation batch is screened.
2. The method for detecting the quality of Schisandra chinensis fermentation based on colony characteristic analysis according to claim 1, characterized in that, S33 involves constructing a source tracing analysis model for the microbial community in the place of origin, quantifying the inherent risk value of the microbial community within each sample analysis unit in the place of origin, specifically including the following steps: S331. Obtain the origin information data and colony distribution data of Citrus aurantium. Take the soil type code and harvest accumulated temperature of Citrus aurantium corresponding to each origin sample analysis unit in the historical monitoring period in the origin information data, and the relative abundance of pathogenic bacteria indicator bacteria of Citrus aurantium in each origin sample analysis unit in the historical monitoring period in the colony distribution data as the classification analysis dataset, and divide the classification analysis dataset into classification analysis training set and classification analysis validation set. S332. Construct a support vector machine classification model. Use the soil type code and harvest accumulated temperature of the trifoliate orange production area corresponding to each production area sample analysis unit in the classification analysis training set as the input features of the support vector machine classification model. Use whether the relative abundance of pathogenic bacteria indicator bacteria of trifoliate orange in each production area sample analysis unit in the classification analysis training set exceeds the safe abundance threshold as the output label of the support vector machine classification model. Train the support vector machine classification model to obtain the initial classification analysis model for each production area sample analysis unit. S333. The initial classification analysis model of each origin sample analysis unit is validated by the classification analysis validation set, and the initial classification analysis model with an accuracy greater than or equal to the preset second model is used as the origin microbial community traceability analysis model for each origin sample analysis unit. S334. Obtain the origin-based microbial community traceability analysis model for each origin-based sample analysis unit; use the decision function distance analysis method to quantify the contribution of each input feature in the origin-based microbial community traceability analysis model of each origin-based sample analysis unit to the classification decision boundary; use the contribution of the input feature as soil type coding in the origin-based microbial community traceability analysis model of each origin-based sample analysis unit as the origin-based inherent risk weight of each origin-based sample analysis unit; multiply the origin-based inherent risk weight of each origin-based sample analysis unit by the relative abundance of pathogenic indicator bacteria of Citrus aurantium in each origin-based sample analysis unit within the current monitoring period to obtain the inherent risk of the microbial community of Citrus aurantium in each origin-based sample analysis unit within the current monitoring period.
3. The method for detecting the quality of Schisandra chinensis fermentation based on colony characteristic analysis according to claim 2, characterized in that, The analysis results of abnormal colony distribution based on the folds of the bitter orange peel and the analysis results of the interference of the bitter orange peel's origin on the fermentation process are used to evaluate the fermentation quality of the current batch of bitter orange peel, including the following specific contents: S41. Obtain the abnormal colony distribution at the folds of the bitter orange peel in each sample analysis unit of the current monitoring period, and at the same time obtain the interference of the bitter orange peel production area of each sample analysis unit of the current monitoring period on the fermentation process. S42. The abnormal colony distribution at the folds of the bitter orange peel in each sample analysis unit of the current production area within the current monitoring period, and the interference of the corresponding bitter orange peel production area on the fermentation process are weighted and summed to obtain the fermentation quality risk of bitter orange peel in each sample analysis unit of the current fermentation batch within the current monitoring period.
4. The method for detecting the fermentation quality of Citrus aurantium based on colony characteristic analysis according to claim 3, characterized in that, The process of screening the bitter orange peel in the current fermentation batch based on the fermentation quality assessment results includes the following specific steps: S51. Extract the fermentation quality risk of Citrus aurantium from each production area sample unit in the current fermentation batch within the current monitoring period; S52. Preset fermentation quality risk threshold, and compare the fermentation quality risk of each sample analysis unit in the current monitoring period with the fermentation quality risk threshold; S53. Discard the bitter orange peel in the corresponding origin sample analysis unit where the fermentation quality risk of bitter orange peel is greater than the fermentation quality risk threshold. S54. Continue fermentation treatment for the bitter orange peel samples in the corresponding origin sample analysis unit whose fermentation quality risk is less than or equal to the fermentation quality risk threshold.
5. A fermentation quality detection system for Citrus aurantium based on colony characteristic analysis, implemented according to any one of claims 1-4, characterized in that, The system includes: The data acquisition module is used to acquire shape characteristic data and origin information data of the current fermentation batch of bitter orange peel, and at the same time acquire colony distribution data during the fermentation process of bitter orange peel; The wrinkled colony detection module is used to analyze the distribution characteristics of colonies during the fermentation process of bitter orange peel by combining the shape feature data and colony distribution data; and to analyze the abnormal colony distribution at the wrinkled position of bitter orange peel based on the analysis results of the colony distribution characteristics. The origin interference detection module is used to combine the origin information data and colony distribution data of Citrus aurantium to analyze the impact of microbial community differences on the fermentation process of Citrus aurantium, and to analyze the interference of Citrus aurantium origin on the fermentation process based on the analysis results of the impact of microbial community differences on the fermentation process of Citrus aurantium. The fermentation quality analysis module is used to evaluate the fermentation quality of the current batch of bitter orange based on the analysis results of abnormal colony distribution at the folds of the bitter orange peel and the analysis results of the interference of the bitter orange peel origin on the fermentation process. The bitter orange peel screening module is used to screen the bitter orange peel in the current fermentation batch based on the fermentation quality assessment results of the current fermentation batch. The control module is used to control the operation of the data acquisition module, the wrinkled colony detection module, the origin interference detection module, the fermentation quality analysis module, and the trifoliate orange screening module.