Intelligent Monitoring Method for the Fermentation Process of Taraxacum chrysantha Tea Based on the Internet of Things

Abstract

This application relates to the field of monitoring and control technologies, and provides an intelligent monitoring method for the fermentation process of Taraxacum officinale var. aurea tea based on the Internet of Things. From the deployment and self-calibration of the sensor array to the extraction and analysis of various data features, and the generation of fermentation stage judgments and control instructions based on this, a precise method logic for fermentation monitoring and control is formed. By obtaining data comprehensively and analyzing it deeply, it is possible to accurately grasp the changes in various parameters during the fermentation process, monitor key indicators such as the mycelial growth rate, the metabolic intensity of the microbial community, and the generation of target products in real time, adjust the fermentation conditions in a timely manner, stabilize the fermentation process at the optimal state, greatly improve the accuracy of fermentation production, reduce the fluctuations in product quality, and ensure the consistency and stability of the quality of each batch of Taraxacum officinale var. aurea tea.

Description

Technical Field

[0001] This application relates to the technical field of monitoring and control, and particularly to an intelligent monitoring method for the fermentation process of Taraxacum chrysantha tea based on the Internet of Things. Background Art

[0002] During the fermentation production process of Taraxacum chrysantha tea, traditional monitoring methods cannot comprehensively obtain various data in the fermentation tank. Moreover, traditional methods rely solely on manual observation or simple image analysis, which are inefficient and have large errors. At the same time, they cannot provide a reliable basis for optimizing the fermentation process. Currently, the judgment of the fermentation stage mainly depends on manual experience, lacking scientific and accurate judgment methods. This makes it impossible to adjust the fermentation strategy in a timely manner according to the actual situation during the fermentation process, easily resulting in over-fermentation or under-fermentation, affecting product quality and output. When regulating the fermentation process, the generation of regulation instructions lacks intelligence and precision, leading to poor regulation effects and increased production costs. Summary of the Invention

[0003] In view of the deficiencies of the prior art, this application provides an intelligent monitoring method for the fermentation process of Taraxacum chrysantha tea based on the Internet of Things. The method includes: deploying a sensor array to respectively obtain environmental data, spectral data, and visual data in the fermentation tank, and self-calibrating the sensor array;

[0004] Extracting the characteristics of the environmental data, spectral data, and visual data to obtain temperature and humidity time series characteristics, gas release rate, spectral characteristics, and hyphal morphology characteristics. Generating the hyphal growth rate according to the temperature and humidity time series characteristics, spectral characteristics, and hyphal morphology characteristics, determining the microbial community metabolism intensity according to the gas release rate and hyphal morphology characteristics, and predicting the generation rate of the target product based on the spectral characteristics;

[0005] Generating a fermentation stage label and confidence according to the hyphal growth rate and microbial community metabolism intensity, and generating a regulation instruction set based on reinforcement learning in combination with the generation rate of the target product;

[0006] Executing the regulation instruction set, monitoring the environmental data to adjust the regulation instruction set and abnormal protection, and synchronizing it to the cloud knowledge base.

[0007] As an optional implementation manner, the hyphal morphology characteristics include hyphal density and hyphal coverage area;

[0008] The extraction logic of the hyphal morphology characteristics includes:

[0009] Preprocessing the hyphal morphology image, and segmenting the hyphal region in the hyphal morphology image according to the gray threshold;

[0010] Skeletonizing the segmented hyphal region through a thinning algorithm to obtain a hyphal skeleton;

[0011] Extract the topological structure of the hyphal skeleton, including the number of branch points, the number of end points, and the average length of hyphal branches;

[0012] Determine the hyphal density and hyphal coverage area based on the topological structure of the hyphal skeleton.

[0013] As an alternative implementation, the generation logic of the hyphal growth rate includes:

[0014] Integrate the temporal characteristics of temperature and humidity, spectral characteristics, and hyphal morphological characteristics at different time points and align them with time stamps;

[0015] Select characteristic points on the hyphal skeleton and record the temporal characteristics of temperature and humidity and spectral characteristics at the positions where each characteristic point is located;

[0016] Calculate the displacement of the characteristic points between two adjacent frames of hyphal morphological images to determine the local growth rate of the characteristic points during this time period, and analyze the changes in the temporal characteristics of temperature and humidity and spectral characteristics during this time period to determine the weights of the characteristic points;

[0017] Obtain the hyphal growth rate by weighted averaging the local growth rates of all characteristic points.

[0018] As an alternative implementation, the determination logic of the metabolic intensity of the microbial community includes:

[0019] Monitor the dynamic changes in the concentrations of carbon dioxide and volatile metabolites, and analyze the gas release rate;

[0020] Perform a correlation analysis between the gas release rate and the hyphal morphological characteristics;

[0021] Quantify the metabolic intensity of the microbial community by weighted averaging based on the release rate of carbon dioxide, the release rate of volatile metabolites, and the change rate of the hyphal coverage area.

[0022] As an alternative implementation, the prediction logic of the production rate of the target product includes:

[0023] Screen the bands of spectral characteristics related to the target product;

[0024] Map the area of the band to the substrate consumption rate, and predict and output the production rate of the target product through a neural network in combination with environmental data;

[0025] Monitor the prediction error of the neural network to determine whether to retrain the neural network.

[0026] As an alternative implementation, the generation logic of the fermentation stage label and confidence includes:

[0027] Integrate the mycelium growth rate, the metabolic intensity of the microbial community, and the historical stage labels to obtain integrated data, and perform clustering analysis on the integrated data based on a clustering algorithm to obtain a clustering result;

[0028] Based on fuzzy logic, divide the mycelium growth rate and the metabolic intensity of the microbial community into fuzzy linguistic variables, and combine the clustering result to determine the fermentation stage labels for different combinations of fuzzy linguistic variables;

[0029] Calculate the confidence level of each fermentation stage label, and correct the fermentation stage label according to the confidence level.

[0030] As an alternative implementation, the generation logic of the regulation instruction set includes:

[0031] Construct a hierarchical reinforcement learning architecture, which includes a goal determination stage and a regulation generation stage;

[0032] Determine the reward function according to the production rate of the target product, the energy consumption, and the pollution risk;

[0033] Monitor the operation data during the fermentation process, and dynamically adjust the regulation generation stage and the reward function.

[0034] As an alternative implementation, the goal determination stage is used to determine the regulation goal according to the fermentation stage label, the confidence level, and the production rate of the target product;

[0035] The regulation generation stage is used to determine the state space and the action space;

[0036] The state space includes temperature, humidity, carbon dioxide concentration, mycelium density, the metabolic intensity of the microbial community, and the production rate of the target product;

[0037] The action space includes heating power, the opening degree of the ventilation valve, and the humidification amount.

[0038] As an alternative implementation, the environmental data includes temperature, humidity, carbon dioxide concentration, and volatile metabolite concentration. The spectral data is used to analyze the metabolic intensity of the microbial community and the production rate of the target product. The visual data includes mycelium morphology images to capture the mycelium growth morphology.

[0039] The deployment logic of the sensor array includes:

[0040] Divide the inside of the fermentation tank into regional blocks according to the structural function of the fermentation tank;

[0041] Within each regional block, deploy the sensor array in a hierarchical three-dimensional manner. The sensor array includes temperature and humidity sensors, gas sensors, spectral sensors, and visual sensors;

[0042] Synchronously activate the sensor array to obtain the environmental data, spectral data, and visual data inside the fermentation tank respectively.

[0043] As an alternative implementation, the self - calibration logic of the sensor array includes:

[0044] Determine the self - calibration conditions according to the changes of time and data, and configure standard data;

[0045] During the self - calibration process, adjust the parameters of the sensor array according to the difference between the acquired data and the standard data;

[0046] Verify the effect of self - calibration after adjustment to feedback whether to perform self - calibration again.

[0047] Compared with the prior art, the beneficial effects of this application are as follows: From the deployment and self - calibration of the sensor array, to the extraction, analysis of various data features, and the generation of fermentation stage judgments and control instructions based on this, a precise fermentation monitoring and control method logic is formed. By obtaining data comprehensively and analyzing it deeply, it can accurately grasp the changes of various parameters during the fermentation process, monitor key indicators such as the hyphal growth rate, the metabolic intensity of the microbial community, and the generation of target products in real - time, timely adjust the fermentation conditions, stabilize the fermentation process at the best state, greatly improve the accuracy of fermentation production, reduce the fluctuation of product quality, and ensure the consistency and stability of the quality of each batch of Taraxacum officinale golden camellia.

[0048] By deploying the sensor array and self - calibrating, comprehensive and accurate environmental data, spectral data, and visual data in the fermentation chamber can be obtained, which lays a solid foundation for the subsequent precise extraction of various data features, ensures the accuracy of data such as temperature and humidity time - series features, gas release rate, spectral features, and hyphal morphological features, and further improves the reliability of calculating the hyphal growth rate, the metabolic intensity of the microbial community, and predicting the generation rate of the target product, providing strong data support for the scientific analysis of the fermentation process.

[0049] Generate fermentation stage labels and confidence levels according to the hyphal growth rate and the metabolic intensity of the microbial community. Compared with traditional manual experience judgment, this method is more scientific and accurate, can reflect the fermentation process in real - time and objectively, provides an important basis for the precise control of the fermentation process, enables operators to adjust the fermentation strategy in a timely manner according to different fermentation stages, avoids over - fermentation or under - fermentation, and effectively improves product quality and yield.

[0050] Combined with the generation rate of the target product, generate a set of control instructions based on reinforcement learning, fully considering various factors such as the generation rate of the target product, energy consumption, and pollution risk. This intelligent control method can achieve precise control of the fermentation process, while ensuring the generation of the target product, reducing energy consumption, reducing pollution risk, improving production efficiency, reducing production costs, and enhancing the market competitiveness of products.

[0051] During the execution of the regulation instruction set, the regulation instruction set is adjusted and abnormal protection is carried out by monitoring environmental data, and it is synchronized to the cloud knowledge base. On the one hand, the regulation strategy can be optimized in a timely manner according to the dynamic changes in the fermentation process to ensure that the fermentation process is always in the best state; on the other hand, the abnormal protection mechanism can quickly detect and handle abnormal situations in the fermentation process to avoid fermentation failure caused by abnormalities. The establishment of the cloud knowledge base is conducive to accumulating data and experience in the fermentation process, providing reference and improvement directions for subsequent fermentation production, and promoting the continuous improvement of the entire fermentation production technology. Brief Description of the Drawings

[0052] To more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts. Among them:

[0053] Figure 1 It is the method flow chart of the intelligent monitoring method for the fermentation process of Taraxacum officinale var. platyphyllum tea based on the Internet of Things provided by the embodiments of the present application;

[0054] Figure 2 It is the logic diagram for extracting the mycelial morphological characteristics of the intelligent monitoring method for the fermentation process of Taraxacum officinale var. platyphyllum tea based on the Internet of Things provided by the embodiments of the present application;

[0055] Figure 3 It is the logic diagram for generating the mycelial growth rate of the intelligent monitoring method for the fermentation process of Taraxacum officinale var. platyphyllum tea based on the Internet of Things provided by the embodiments of the present application. Detailed Embodiments

[0056] To make the objectives, technical solutions, and advantages of the embodiments of the present application more obvious and understandable, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the specification. Obviously, the described embodiments are only some embodiments of the present application, rather than all embodiments.

[0057] Embodiment

[0058] The application scenario of the embodiments of the present application is to produce a more effective probiotic "double bacteria" Taraxacum officinale var. platyphyllum tea through microbial fermentation. The base tea is made by taking young leaves and going through processes such as picking, washing, steaming, sun-drying, rolling, frying, and drying. Later, through modern microbial technologies, a series of fungal cultivation and tea-making technologies such as sterilization, inoculation, cultivation, drying, and shaping are carried out. Eurotium amstelodami is precisely placed, and the optimal fermentation environment is precisely regulated to produce a high amount of Eurotium cristatum (golden flower bacteria). The content of taraxasterol in Taraxacum officinale is increased through microbial fermentation to produce a more effective probiotic "double bacteria" Taraxacum officinale var. platyphyllum tea.

[0059] As shown Figure 1 in the figure, the method flowchart of the intelligent monitoring method for the fermentation process of Taraxacum officinale flower tea based on the Internet of Things provided by the embodiment of the present application is shown. The method includes:

[0060] S1. Deploy a sensor array to respectively obtain the environmental data, spectral data and visual data in the fermentation tank, and self-calibrate the sensor array.

[0061] The environmental data includes temperature, humidity, carbon dioxide concentration and volatile metabolite concentration. The spectral data is used to analyze the metabolic intensity of the flora and the generation rate of the target product. The visual data includes hyphal morphology images to capture the growth morphology of the hyphae.

[0062] The deployment logic of the sensor array includes:

[0063] Divide the inside of the fermentation tank into regional blocks according to the structural function of the fermentation tank;

[0064] In each regional block, deploy the sensor array in a hierarchical three-dimensional manner. The sensor array includes temperature and humidity sensors, gas sensors, spectral sensors and visual sensors;

[0065] Synchronously activate the sensor array to respectively obtain the environmental data, spectral data and visual data in the fermentation tank.

[0066] There are differences in the fermentation conditions at different positions in the fermentation tank. For example, the temperature, humidity and gas concentration change differently in the areas near the heating equipment or ventilation openings. Dividing the regional blocks can more accurately monitor the fermentation environment in different regions and provide a basis for subsequent precise regulation; through on-site investigation of the fermentation tank, combined with the internal layout of the fermentation tank, such as the location of the fermentation equipment and the direction of the ventilation duct, etc., digitally model the fermentation tank through 3D modeling software. According to the model of the fermentation tank, divide the fermentation tank into several regional blocks of appropriate size. For example, for a fermentation tank with a length of 10 meters, a width of 8 meters and a height of 5 meters, it can be divided into 8 regional blocks of 2.5 meters × 4 meters × 5 meters to ensure that the environmental characteristics of each regional block have a certain representativeness during the fermentation process; thus realizing the refined monitoring of the environment in the fermentation tank by region, improving the accuracy and pertinence of data collection, and providing a clear position framework for the hierarchical three-dimensional deployment of sensors, making the deployment of the sensor array more planned and better able to obtain the environmental data of each region.

[0067] The environmental conditions at different heights inside the fermentation tank also vary. The layered three-dimensional deployment can comprehensively obtain the environmental data at different height levels inside the fermentation tank, more truly reflecting the spatial changes in the fermentation environment. In each regional block, along the height direction of the fermentation tank, a sensor plane is set every 1 meter. On each plane, various sensors are evenly distributed at a certain interval. For example, temperature and humidity sensors are used to obtain the temperature and humidity inside the fermentation tank, one is installed every 2 meters. Gas sensors include carbon dioxide sensors and volatile metabolite sensors, which are used to monitor the carbon dioxide concentration and volatile metabolite concentration respectively, with each sensor spaced 3 meters apart. Spectral sensors are used for subsequent analysis of the metabolic intensity of the microbial community and the production rate of the target product, and one is installed at the center position of each regional block. Visual sensors need to be equipped with lenses that can adapt to the light conditions inside the fermentation tank and are installed on adjustable-angle brackets to capture images of the mycelial morphology. Two are installed in each regional block, taking pictures from different angles respectively. Thus, more comprehensive and accurate fermentation environment data can be obtained, providing multi-dimensional information for analyzing the fermentation process, and ensuring that when the sensor array is synchronously activated, rich environmental data, spectral data, and visual data can be obtained from different heights and positions, providing sufficient data support for subsequent data analysis and fermentation process control.

[0068] To ensure that each sensor obtains data at the same moment, guarantee the time synchronization of the data, and make the subsequent data analysis more reliable and effective; set a unified trigger time for each sensor, send a synchronous trigger command to the sensors in each regional block, so that the temperature and humidity sensors, gas sensors, spectral sensors, and visual sensors start obtaining data simultaneously. Thus, it is ensured that the obtained environmental data, spectral data, and visual data have time consistency, avoiding analysis errors caused by asynchronous data acquisition times, and providing a reliable data basis for accurately analyzing the relationship between environmental factors, microbial community metabolism, and mycelial growth during the fermentation process, which is conducive to precise fermentation control based on the data.

[0069] The self-calibration logic of the sensor array includes:

[0070] Determine the self-calibration conditions according to the changes in time and data, and configure standard data;

[0071] During the self-calibration process, adjust the parameters of the sensor array according to the difference between the obtained data and the standard data;

[0072] Verify the effect of the self-calibration after adjustment to feedback whether to perform self-calibration again.

[0073] During long-term use, sensors are affected by environmental factors (such as temperature and humidity changes) or their own aging, resulting in deviations in the acquired data. Setting self-calibration conditions and configuring standard data can promptly detect and correct these deviations, ensuring the accuracy of the data acquired by the sensors. The self-calibration condition is set to perform regular calibration every 24 hours. At the same time, when the data acquired by the sensor shows abnormal fluctuations or exceeds the preset threshold range, self-calibration is immediately triggered. The situation of abnormal fluctuations refers to, for example, the temperature data changing by more than ±5°C within 10 minutes, and there is no similar change in the temperature sensors in other area blocks. The situation of exceeding the preset threshold range refers to, for example, the carbon dioxide concentration exceeding ±20% of the normal fermentation range. The standard data is obtained by measuring a high-precision standard source in a laboratory environment. For example, a thermo-hygrostat is used as the standard data for temperature and humidity, with a temperature accuracy of up to ±0.1°C and a humidity accuracy of up to ±2%RH. A standard gas generator is used to configure carbon dioxide and volatile metabolite standard gases with known concentrations for calibrating gas sensors. The standard data for spectral sensors is obtained by measuring standard samples with known spectral characteristics. These standard data are stored in a database for use during self-calibration. Thus, it is possible to promptly detect and correct the acquisition deviations of the sensors, ensuring the accuracy and reliability of the sensor data, and providing an accurate comparison benchmark for subsequent difference calculations between the acquired data and the standard data, making the adjustment of sensor array parameters more targeted.

[0074] By comparing the difference between the data acquired by the sensor and the standard data, the sensor parameters are adjusted to make the result of the data acquired by the sensor closer to the true value, ensuring the accuracy of the data acquired by the sensor. During the self-calibration process, the sensor acquires data again and compares the acquired data with the standard data. For thermo-hygro sensors, if the difference between the acquired temperature value and the standard temperature value is greater than the allowable error range, such as ±0.5°C, the temperature compensation coefficient of the thermo-hygro sensor is adjusted according to the size of the difference. For gas sensors, if the difference between the acquired carbon dioxide concentration and the standard concentration is too large, the sensitivity parameter of the gas sensor is adjusted for calibration. For spectral sensors, according to the difference between the acquired spectral data and the standard spectrum, the wavelength calibration parameter and the intensity calibration parameter are adjusted. Thus, the measurement accuracy of the sensor is effectively improved, ensuring the accuracy of the data. Moreover, for the sensor after parameter adjustment, the acquired data is more accurate, providing reliable data for verifying the self-calibration effect and facilitating the judgment of whether the self-calibration is successful.

[0075] Determine whether the self-calibration has achieved the expected effect. If not, perform self-calibration again to ensure that the sensor is always in the best working state. After the sensor completes the parameter adjustment, obtain the data again and compare it with the standard data to calculate the error between the adjusted data and the standard data. If the error is within the allowable error range, it is determined that the self-calibration is successful. If the error is still greater than the allowable error range, re-execute the self-calibration process. At the same time, record the results of each self-calibration in the database, including the calibration time, error data before and after calibration, etc., for subsequent analysis of the performance change trend of the sensor. Thus, ensure the effectiveness of the sensor self-calibration, ensure that the sensor continuously and stably provides accurate data. If the self-calibration is successful, the sensor can continue to work normally and provide reliable data for the fermentation process monitoring. If the self-calibration fails and is re-performed, it can avoid fermentation control mistakes caused by inaccurate sensor data and ensure the smooth progress of the fermentation process.

[0076] S2. Extract the features of environmental data, spectral data, and visual data to obtain the temperature and humidity time series features, gas release rate, spectral features, and hyphal morphological features. Generate the hyphal growth rate based on the temperature and humidity time series features, spectral features, and hyphal morphological features. Determine the microbial community metabolic intensity based on the gas release rate and hyphal morphological features, and predict the production rate of the target product based on the spectral features.

[0077] The hyphal morphological features include hyphal density and hyphal coverage area;

[0078] As Figure 2 shown, the extraction logic of the hyphal morphological features includes:

[0079] Preprocess the hyphal morphological image and segment the hyphal region in the hyphal morphological image according to the gray threshold;

[0080] Skeletonize the segmented hyphal region through a thinning algorithm to obtain the hyphal skeleton;

[0081] Extract the topological structure of the hyphal skeleton, including the number of branch points, the number of end points, and the average length of hyphal branches;

[0082] Determine the hyphal density and hyphal coverage area according to the topological structure of the hyphal skeleton.

[0083] The original hyphal morphological images have problems such as noise and uneven illumination, which will affect the accuracy of subsequent feature extraction. By preprocessing to remove noise and segmenting the hyphal region according to the gray threshold, the hyphal morphology can be highlighted, facilitating subsequent analysis. The Gaussian filtering algorithm is used to denoise the hyphal morphological images. The Gaussian filtering algorithm effectively smooths the images and reduces noise interference by performing weighted averaging on the neighborhood of each pixel point in the hyphal morphological images. Then, the Otsu algorithm automatically calculates an appropriate gray threshold. The Otsu algorithm is based on the gray histogram of the image and determines the optimal gray threshold by maximizing the between-class variance. The part of the hyphal morphological image with a gray value greater than or equal to the gray threshold is identified as the hyphal region, and the part less than the gray threshold is removed as the background, thus achieving accurate segmentation of the hyphal region. This improves the image quality, accurately separates the hyphal region, provides a clear image basis for subsequent feature extraction, and the segmented hyphal region provides an accurate object for skeletonization processing, ensuring that the subsequent thinning algorithm can accurately act on the hyphae and avoiding background interference.

[0084] The shape of the hyphae is relatively complex. The hyphal skeleton obtained through skeletonization processing can simplify the shape of the hyphae, more clearly display the growth structure and branching direction of the hyphae, and facilitate the extraction of topological structure features. Through a morphological thinning algorithm, by continuously deleting edge pixels and retaining skeleton pixels, the hyphal region is gradually thinned into a single-pixel-wide skeleton. During the processing, according to specific deletion rules, the edge pixels of the hyphal region are iteratively deleted until no more deletion is possible, and finally a complete hyphal skeleton is obtained. This simplifies the shape of the hyphae, highlights the growth structure and branching characteristics of the hyphae, provides convenience for topological structure extraction, and the generated hyphal skeleton is the basis for extracting topological structure features. An accurate hyphal skeleton can ensure the accuracy and reliability of topological features such as the number of branch points and the number of end points extracted.

[0085] The topological structure features of the hyphal skeleton can reflect the growth pattern and complexity of the hyphae, which is of great significance for analyzing the growth state and metabolic situation of the hyphae. Traverse each pixel point of the hyphal skeleton, and identify branch points and end points by judging the neighborhood pixel conditions of the pixel points. For example, when there are three or more pixel points in the neighborhood of a pixel point connected to it, the pixel point is identified as a branch point; when there is only one pixel point in the neighborhood of a pixel point connected to it, the pixel point is identified as an end point. For the average length of the hyphal branches, it can be obtained by tracking the path length from the branch point to the end point and averaging all branch lengths. This enables the acquisition of key topological features that can reflect the growth state of the hyphae, providing a quantitative basis for in-depth analysis of hyphal growth. The extracted topological structure features are an important basis for determining the hyphal density and hyphal coverage area, directly affecting the subsequent comprehensive evaluation of the hyphal growth state.

[0086] Hyphal density and hyphal coverage area are important indicators for measuring the growth scale and vitality of hyphae, and are of great value for evaluating the growth status and fermentation effect of hyphae during fermentation. Among them, hyphal density is determined by calculating the ratio of the number of pixel points of the hyphal skeleton to the total area of the hyphal morphology image. For the hyphal coverage area, first binarize the segmented hyphal region, then count the number of pixel points in the hyphal region of the binarized image, and then convert the number of pixels into the actual coverage area according to the conversion relationship between the pixels of the hyphal morphology image and the actual area. Thus, the growth scale and vitality of hyphae are quantified, providing an intuitive quantitative index for evaluating the fermentation effect. As important components of hyphal morphological characteristics, hyphal density and hyphal coverage area provide key data support for determining the metabolic intensity of the microbial community and analyzing the hyphal growth rate.

[0087] As Figure 3 shown, the generation logic of the hyphal growth rate includes:

[0088] Integrate the temperature and humidity time series characteristics, spectral characteristics, and hyphal morphological characteristics at different time points and align them with time stamps;

[0089] Select feature points on the hyphal skeleton and record the temperature and humidity time series characteristics and spectral characteristics at the location of each feature point;

[0090] Calculate the displacement of the feature points between two adjacent frames of hyphal morphology images to determine the local growth rate of the feature points during this time period, and analyze the changes in the temperature and humidity time series characteristics and spectral characteristics during this time period to determine the weight of the feature points;

[0091] Obtain the hyphal growth rate by weighted averaging the local growth rates of all feature points.

[0092] The temperature and humidity time series characteristics, spectral characteristics, and hyphal morphological characteristics reflect the fermentation environment and hyphal growth conditions from different angles. Integrating them and aligning them with time stamps can comprehensively consider the influence of various factors on hyphal growth and ensure the accuracy of subsequent analysis. Store the temperature and humidity time series data, spectral data, and hyphal morphological characteristic data obtained at different time points in chronological order, and ensure the accurate association of different types of data at the same time point through timestamp matching. For example, each time data is obtained, record the timestamp accurate to seconds. During data fusion, integrate the corresponding data according to the timestamp. Thus, a comprehensive and time-synchronized data foundation is provided, creating conditions for accurately analyzing the relationship between the hyphal growth rate and various factors. Moreover, when it is convenient to select feature points on the hyphal skeleton after integration, the corresponding multi-faceted information of the feature points can be obtained simultaneously, providing rich data support for subsequent analysis.

[0093] Selecting representative feature points on the hyphal skeleton and recording their related temperature, humidity, and spectral characteristics helps analyze the local growth of hyphae under different environmental conditions, and further determine the influence weights of various factors on hyphal growth; uniformly select multiple feature points on the hyphal skeleton, and the interval can be determined according to the hyphal density and image resolution. For example, select a feature point every 10 pixels. For each selected feature point, find the temporal characteristics of temperature and humidity and spectral characteristics at its corresponding time point, and associate and store this information with the feature point; thus, obtain multi-feature information of feature points at different positions, providing specific data samples for analyzing the local growth rate and the weights of various factors, directly affecting the accuracy of subsequent hyphal growth rate calculation.

[0094] Calculating the displacement of a feature point between two adjacent frames of images to obtain the local growth rate, and analyzing the changes in temperature, humidity, and spectral characteristics to determine the weights can comprehensively consider the influence of environmental factors on the local growth of hyphae, making the calculated hyphal growth rate more in line with the actual situation; by comparing the coordinates of feature points in two adjacent frames of hyphal morphology images and using the Euclidean distance formula to calculate the displacement of the feature point. For example, the coordinates of a feature point in the first frame of the image are , and the coordinates in the second frame of the image are , then the displacement of the feature point between two adjacent frames of images is . Combining the time interval obtained from the two frames of images, the moving distance of the feature point per unit time, that is, the local growth rate, is obtained. For example, a certain feature point has a displacement of 10 pixels between images collected in two adjacent hours. Given that the conversion ratio between image pixels and actual length is 1 pixel = 0.1 mm and the time interval is 2 hours, the local growth rate at this feature point is 0.5 mm / h.

[0095] For determining the weights, for the temporal characteristics of temperature and humidity, taking the temperature and humidity values of each feature point at the first time point as the benchmark, divide the temperature and humidity values of subsequent time points by this benchmark value to obtain a dimensionless temperature and humidity feature sequence, and synchronously process the spectral characteristics, thereby eliminating the difference in data dimensions, enabling different types of data to be analyzed on the same scale, and improving the accuracy and reliability of grey relational analysis; let the local growth rate data sequence be the reference sequence, and the temperature and humidity feature sequence and the spectral feature sequence be the comparison sequences. Calculate the absolute difference between the reference sequence and the comparison sequences of each feature point at each time point, then determine the two-level minimum difference and the two-level maximum difference, and finally calculate the correlation coefficient according to the grey correlation coefficient formula to quantify the degree of association between each factor and the local growth rate at each feature point at different time points, providing specific data support for subsequent comprehensive calculation of the correlation degree.

[0096] Arithmetically average the grey correlation coefficients of each feature point to obtain the correlation degrees between the temperature and humidity time series features and spectral features of the feature point and the local growth rate. For example, for a certain feature point, calculate the correlation degree between its temperature and humidity time series features and the local growth rate and the correlation degree between its spectral features and the local growth rate, so as to comprehensively consider the correlation situations at each time point and obtain a quantitative index that can reflect the influence degree of each factor on the local growth of hyphae, providing a clear basis for weight determination; the higher the correlation degree, the greater the influence of the factor on the local growth of hyphae. Therefore, a higher weight is assigned to it, which can more reasonably comprehensively consider the influence of each factor on the hyphal growth rate and make the calculated hyphal growth rate more in line with the actual situation; for each feature point, normalize the correlation degrees of the temperature and humidity time series features and spectral features to obtain the weights of each factor, so as to reasonably distribute the weights of each factor, fully consider the influence of each factor when calculating the hyphal growth rate, and improve the accuracy and scientificity of the hyphal growth rate calculation; thus, comprehensively consider environmental factors, accurately calculate the local growth rate, and determine the weights of each factor, making the calculation of the hyphal growth rate more scientific. The obtained local growth rate and the weights of each factor are the key data for calculating the overall hyphal growth rate and directly determine the accuracy of the final hyphal growth rate.

[0097] Obtain the hyphal growth rate by weighted averaging the local growth rates of all feature points, which can comprehensively reflect the overall growth situation of hyphae and provide a quantitative comprehensive index for evaluating the growth state of hyphae during fermentation; according to the local growth rates and corresponding weights of each feature point determined previously, use the weighted average formula to calculate the hyphal growth rate; thus, provide a quantitative index that can comprehensively reflect the overall growth situation of hyphae, which is convenient for intuitively evaluating the growth state of hyphae and provides an important basis for determining the metabolic intensity of the microbial community and adjusting the fermentation conditions.

[0098] The determination logic of the microbial community metabolic intensity includes:

[0099] Monitor the dynamic changes in the carbon dioxide concentration and the concentration of volatile metabolites, and analyze the gas release rate;

[0100] Conduct a correlation analysis between the gas release rate and the hyphal morphological features;

[0101] Quantitatively weight and obtain the microbial community metabolic intensity according to the release rate of carbon dioxide, the release rate of volatile metabolites, and the change rate of the hyphal coverage area.

[0102] The dynamic changes in carbon dioxide concentration and volatile metabolite concentration can directly reflect the metabolic activities of the microbial community. Analyzing their release rates can quantify the activity level of microbial metabolism. By using gas sensors, such as carbon dioxide sensors and volatile metabolite sensors, the gas concentration changes in the fermentation chamber are monitored in real time. Data is obtained at regular time intervals (e.g., every 10 minutes). By calculating the ratio of the difference between two adjacent data acquisitions to the time interval, the gas release rate is obtained. For example, if the carbon dioxide concentration increases from 3% to 3.5% within 10 minutes, the carbon dioxide release rate is (3.5% - 3%) ÷ 10 = 0.05% / minute. Thus, it can reflect the intensity of microbial metabolic activities in real time and quantitatively, providing key data for subsequent analysis. Moreover, accurate gas release rate data is the basis for correlation analysis with hyphal morphological characteristics and affects the accuracy of determining the microbial metabolic intensity.

[0103] Microbial metabolism is closely related to hyphal growth. Correlating the gas release rate with hyphal morphological characteristics for analysis can provide a more comprehensive understanding of the actual situation of microbial metabolism and serve as a basis for accurately determining the microbial metabolic intensity. Through correlation analysis methods, such as Pearson correlation coefficient analysis, the correlations between the carbon dioxide release rate, volatile metabolite release rate and hyphal morphological characteristics such as hyphal density and hyphal coverage area are calculated. Through analysis, it is found that when the hyphal coverage area increases, the carbon dioxide release rate usually also increases, indicating a positive correlation between the two. Thus, it reveals the internal relationship between microbial metabolism and hyphal growth, provides a more comprehensive perspective for quantifying the microbial metabolic intensity, and the results of the correlation analysis provide a reference for determining the weights of various factors in quantifying the microbial metabolic intensity, affecting the accuracy of the final calculation of the microbial metabolic intensity.

[0104] By weighted quantification of the carbon dioxide release rate, volatile metabolite release rate and hyphal coverage area change rate, multiple factors can be comprehensively considered to obtain a quantification index that can accurately reflect the microbial metabolic intensity. According to the correlation results obtained from the correlation analysis, the weights of various factors are determined. For example, after analysis, it is found that the carbon dioxide release rate has a greater impact on the microbial metabolic intensity, and a weight of 0.4 is assigned to it; the weight of the volatile metabolite release rate is 0.3; the weight of the hyphal coverage area change rate is 0.3. Then, the microbial metabolic intensity is calculated by weighted calculation. Thus, a comprehensive quantification index of the microbial metabolic intensity is provided, which intuitively reflects the metabolic activity level of the microbial community during fermentation and provides an important basis for judging whether the fermentation state is normal and whether fermentation conditions need to be adjusted.

[0105] The prediction logic of the production rate of the target product includes:

[0106] Screening the spectral characteristic bands related to the target product;

[0107] Map the area of the spectral band to the substrate consumption rate, and combine environmental data to predict the production rate of the target product through a neural network;

[0108] Monitor the prediction error of the neural network to determine whether to retrain the neural network.

[0109] Spectral data contains a large amount of information, but not all spectral bands are related to the production of the target product. Screening out the spectral bands related to the target product can reduce data redundancy and improve the accuracy and efficiency of prediction; through the least squares discriminant analysis method, the spectral data is analyzed. This method establishes a regression model between the spectral data and the concentration of the target product, and finds out the spectral bands that have the greatest impact on the change of the target product concentration. For example, when analyzing the spectral data of the fermentation process of Taraxacum chrysantha tea, it is found that the spectral bands in the range of 700 - 800 nm and 1200 - 1300 nm are closely related to the production of the target product; thus reducing the amount of data processing, highlighting the key information related to the production of the target product, improving the accuracy and efficiency of the prediction model. The screened spectral feature bands are the basis for subsequent mapping the substrate consumption rate and predicting the production rate of the target product, directly affecting the prediction accuracy.

[0110] Mapping the area of the spectral bands with spectral features to the substrate consumption rate, and combining environmental data to predict the production rate of the target product through a neural network can comprehensively consider the influence of spectral information and environmental factors on the production of the target product and achieve an effective prediction of the production rate of the target product; through experiments, it is found that there is a linear relationship between the area of specific spectral bands and the substrate consumption rate, and a linear regression equation can be used for fitting. The obtained substrate consumption rate and environmental data (such as temperature, humidity, and oxygen concentration, etc.) are used as inputs and input into a pre-trained neural network model. The neural network model can adopt a multi-layer perceptron, and the model is trained with a large amount of experimental data to enable it to learn the complex relationship between the input data and the production rate of the target product, thereby predicting and outputting the production rate of the target product; thus comprehensively considering various factors, achieving an effective prediction of the production rate of the target product, providing a basis for the control and optimization of the fermentation process. The predicted production rate of the target product is an important reference for judging the fermentation effect and adjusting the fermentation conditions, and at the same time, the prediction error is also the basis for judging whether to retrain the neural network.

[0111] As the fermentation process progresses, the environment and fermentation conditions change, leading to a decline in the prediction accuracy of the neural network. By monitoring the prediction error, detecting the inaccuracy of the neural network model in a timely manner, and retraining the neural network, the prediction accuracy can be ensured; after predicting the production rate of the target product each time, the predicted value is compared with the actual measured value, and the prediction error is calculated. For example, the mean square error is used as the error index. When the prediction error is greater than the preset threshold (such as 5%), new experimental data is collected, the neural network is retrained, and the parameters of the neural network model are updated to improve the prediction accuracy; thus ensuring the accuracy of the prediction of the production rate of the target product, enabling the prediction results to better guide the control and optimization of the fermentation process. The accurate prediction results provide a reliable basis for adjusting the fermentation conditions and optimizing the fermentation process, and contribute to improving the fermentation efficiency and product quality.

[0112] S3. Generate fermentation stage labels and confidence levels based on the hyphal growth rate and the microbial community metabolic intensity, and generate a regulation instruction set based on reinforcement learning in combination with the production rate of the target product.

[0113] The generation logic of the fermentation stage labels and confidence levels includes:

[0114] Integrate the hyphal growth rate, the microbial community metabolic intensity, and the historical stage labels to obtain integrated data, and perform clustering analysis on the integrated data based on the clustering algorithm to obtain the clustering results;

[0115] Divide the hyphal growth rate and the microbial community metabolic intensity into fuzzy linguistic variables based on fuzzy logic, and determine the fermentation stage labels for different combinations of fuzzy linguistic variables in combination with the clustering results;

[0116] Calculate the confidence level for each fermentation stage label, and correct the fermentation stage label according to the confidence level.

[0117] The mycelial growth rate, the metabolic intensity of the microbial community, and the historical stage labels each contain partial information about the fermentation process. Integrating these data and performing clustering analysis can mine data features from multiple dimensions, discover potential patterns in different fermentation stages, and provide a basis for accurately dividing the fermentation stages. Integrate the real-time obtained mycelial growth rate and the metabolic intensity of the microbial community with the historical stage label data stored in the database. For example, record the mycelial growth rate and the metabolic intensity of the microbial community every 2 hours, and at the same time associate the corresponding historical stage labels. Perform clustering analysis through the density-based spatial clustering algorithm. First, set appropriate neighborhood radii (such as eps = 0.5) and minimum sample numbers (such as minPts = 5) according to the data characteristics. The density-based spatial clustering algorithm will scan the data points, divide the density-connected data points into the same cluster, and at the same time identify the outliers. Divide the integrated data into multiple clusters through the density-based spatial clustering algorithm, and each cluster represents a fermentation state with similar characteristics. Extract valuable information from multi-dimensional data, preliminarily classify different states in the fermentation process, and lay a foundation for accurately determining the fermentation stage labels subsequently. The clustering results provide a reference framework for the fuzzy logic to divide the fermentation stage labels, making the application of fuzzy logic more targeted and reasonable.

[0118] The changes in the mycelial growth rate and the metabolic intensity of the microbial community during the fermentation process are continuous, and it is difficult to divide the fermentation stages with precise numerical boundaries. Fuzzy logic can transform these continuously changing variables into fuzzy linguistic variables, which is more in line with the fuzzy characteristics of the stage division in the fermentation process. Combining the clustering results can more accurately determine the fermentation stage labels. The fermentation stage labels include the germination period, the growth period, the stable period, and the decline period. Divide the mycelial growth rate into three fuzzy linguistic variables: "low", "medium", and "high", and the metabolic intensity of the microbial community is also divided into three fuzzy linguistic variables: "weak", "medium", and "strong". For example, when the mycelial growth rate is less than a certain threshold (such as 0.1 mm / h), it is determined as "low", in the threshold range (0.1 - 0.5 mm / h) it is "medium", and greater than 0.5 mm / h it is "high". The metabolic intensity of the microbial community is comprehensively judged according to indicators such as the carbon dioxide release rate and the volatile metabolite release rate. Combining the clustering results, formulate a fuzzy rule table. For example, when the mycelial growth rate is "high" and the metabolic intensity of the microbial community is "strong", and at the same time the clustering results show that the characteristics of this group of data are most similar to those of the "growth period" in the historical data, determine the fermentation stage label of this stage as "growth period". Thus, it fully considers the fuzziness of the variables in the fermentation process, is more in line with the actual fermentation situation, improves the accuracy and reasonableness of the fermentation stage division, and the clear fermentation stage labels provide a basis for calculating the confidence level and correcting the labels, and at the same time also provide key information for determining the subsequent control objectives.

[0119] The fermentation process is affected by various factors and is uncertain. Calculating the confidence level can quantify the reliability of the judgment of the fermentation stage label. When the confidence level is low, correcting the label can improve the accuracy of the label and provide a more reliable basis for subsequent regulation; calculating the confidence level through the evidence theory. For example, taking the fermentation stage label determined by fuzzy logic as evidence, calculating the confidence level of each label according to the conflict degree and support degree between different evidences. Suppose the confidence level of the "growth period" label is calculated to be 0.75 (less than 0.85). At this time, the label needs to be corrected. By re-examining the change trends of the mycelial growth rate and the microbial community metabolism intensity, as well as the comparison with historical data, if it is found that although the microbial community metabolism intensity is in the "strong" range, there is a downward trend, and combined with some characteristics similar to the "stable period" in the clustering results, the fermentation stage label is corrected to the "stable period"; thus improving the reliability of the fermentation stage label, reducing the wrong judgment caused by uncertainty, making the subsequent regulation decision based on the label more scientific, and the corrected fermentation stage label and confidence level provide more accurate input for the target determination stage, affecting the determination of the regulation target and the formulation of the regulation strategy.

[0120] The generation logic of the regulation instruction set includes:

[0121] Construct a hierarchical reinforcement learning architecture, which includes a target determination stage and a regulation generation stage;

[0122] Determine the reward function according to the production rate, energy consumption, and pollution risk of the target product;

[0123] Monitor the operation data during the fermentation process and dynamically adjust the regulation generation stage and the reward function.

[0124] The fermentation process is complex, involving multiple goals and various regulation means. The hierarchical reinforcement learning architecture can decompose the complex regulation task into subtasks at different levels, making the learning process more efficient, facilitating the separate handling of the problems of target determination and regulation generation, and improving the generation quality of the regulation instruction set; through a two-layer architecture, the upper layer is the target determination stage, and the lower layer is the regulation generation stage. In the target determination stage, determine the regulation target according to the fermentation stage label, confidence level, and the production rate of the target product; in the regulation generation stage, generate specific regulation instructions according to the determined regulation target, combined with the state space and the action space. Implement the hierarchical architecture through programming, use the reinforcement learning library of Python to build a framework, and define the input, output, and processing logic of different stages; thus decomposing the complex regulation problem, improving the efficiency and accuracy of reinforcement learning, making the generation of the regulation instruction set more targeted and systematic, providing a structured framework for subsequent determination of the reward function, monitoring operation data, and dynamic adjustment, and ensuring the orderly progress of the regulation instruction set generation process.

[0125] The reward function is the core of reinforcement learning and is used to guide the agent (the regulation system in the application scenario of this embodiment) to learn the optimal policy. By determining the reward function based on the production rate of the target product, energy consumption, and pollution risk, the regulation system can pursue the production of the target product while taking into account energy consumption and pollution control, maximizing the comprehensive benefits. The reward function is defined as , where is the production rate of the target product, is the energy consumption, is the pollution risk, , and are weight coefficients, which are determined through experiments and data analysis. For example , and ; the higher the production rate of the target product, the higher the reward, and the higher the energy consumption and pollution risk, the lower the reward. For example, when the production rate of the target product increases by 10%, the reward increases by 0.6, and when the energy consumption increases by 10%, the reward decreases by 0.2. Thus, it guides the regulation system to balance multiple objectives, optimize the fermentation process, improve production efficiency, reduce costs and environmental impacts. The reward function determines the learning direction of the regulation system, affects the utilization of the state space and action space in the regulation generation stage, and the generation of the final regulation instruction set.

[0126] The fermentation process is dynamically changing, and the initially set regulation strategy may not be able to adapt to various changes in the process. Monitoring the operation data during the fermentation process and dynamically adjusting the regulation generation stage and the reward function can enable the regulation system to respond to the changes in the fermentation process in real time, maintaining the effectiveness and adaptability of the regulation; obtain the operation data during the fermentation process every certain period (such as 30 minutes), including temperature, humidity, carbon dioxide concentration, hyphal density, microbial metabolism intensity, and the production rate of the target product, etc., analyze the change trends of these data. If it is found that the production rate of the target product is lower than expected for a long time and the energy consumption is too high, it indicates that the current regulation strategy is unreasonable. At this time, adjust the weight coefficients of the reward function, for example, appropriately increase , reduce , to encourage the regulation system to pay more attention to improving the production rate of the target product, and at the same time re-evaluate the state space and action space according to the data changes. For example, if it is found that the temperature change has a greater impact on the production of the target product, the quantization interval of the temperature in the state space can be appropriately reduced to improve the accuracy of temperature regulation; thus enabling the regulation system to adapt to the dynamic changes of the fermentation process, maintaining good regulation effects, improving the quality and yield of the fermentation products. The dynamically adjusted regulation generation stage and reward function provide more realistic parameters and conditions for the target determination stage and the generation of the regulation instruction set, ensuring the accuracy and effectiveness of the regulation instructions.

[0127] The target determination stage is used to determine the regulation target according to the fermentation stage label, confidence level, and the production rate of the target product;

[0128] The regulation generation stage is used to determine the state space and action space;

[0129] The state space includes temperature, humidity, carbon dioxide concentration, hyphal density, microbial community metabolic intensity, and the production rate of the target product;

[0130] The action space includes heating power, ventilation valve opening degree, and humidification amount.

[0131] The fermentation stage label, confidence level, and the production rate of the target product reflect the current fermentation state and expected goal. By comprehensively considering this information to determine the regulation target, the regulation strategy can be made more targeted to meet the requirements of different fermentation stages; when the fermentation stage label is "growth period" and the confidence level is relatively high (such as 0.9), and at the same time the production rate of the target product is lower than expected, the regulation target is determined as "to increase the production rate of the target product on the premise of ensuring the stability of the microbial community metabolic intensity"; if the fermentation stage label is "stable period", but the production rate of the target product begins to decline and the confidence level is 0.8, the regulation target can be set as "to maintain the current production rate of the target product and control the energy consumption and pollution risk". By writing a logical judgment program, according to different combinations of fermentation stage labels, confidence levels, and production rates of the target product, the corresponding regulation targets are determined and stored in the target library of the regulation system; thus, it clarifies the specific regulation direction for the regulation system, makes the regulation more targeted, improves the controllability of the fermentation process, and the determined regulation target provides guidance for the regulation generation stage, affecting the selection of the state space and action space and the generation of regulation instructions.

[0132] Defining the state space and action space is the basis for the reinforcement learning to generate control instructions. The state space covers the key factors affecting the fermentation process, and the action space includes the operations available for regulating the fermentation process. Reasonably determining these two spaces enables the control system to search for the optimal control strategy within a feasible range. Among them, the state space includes temperature (set range: 20 - 35 °C, quantization interval: 1 °C), humidity (set range: 50% - 80%, quantization interval: 5%), carbon dioxide concentration (set range: 0.1% - 0.5%, quantization interval: 0.05%), hyphal density, microbial community metabolic intensity, and the production rate of the target product. The action space includes heating power (set range: 0 - 10 kW, adjustment interval: 0.5 kW), ventilation valve opening (set range: 0 - 100%, adjustment interval: 5%), and humidification amount (set range: 0 - 5 L / h, adjustment interval: 0.5 L / h). Store this state and action information in the control system in the form of a data structure for easy processing by the reinforcement learning algorithm. Thus, it provides a clear control range and operation options for the control system, enabling the reinforcement learning to search for the optimal control strategy within a reasonable range, improving the accuracy and effectiveness of control. The determined state space and action space are the basis for the reinforcement learning algorithm to generate a control instruction set, which determines the specific form and value range of the control instructions.

[0133] S4. Execute the control instruction set, monitor the environmental data to adjust the control instruction set and abnormal protection, and synchronize it to the cloud knowledge base.

[0134] The fermentation process is dynamically changing. Obtaining and analyzing environmental data in real time can promptly grasp the fermentation situation, determine whether the current control instruction set is effective, and provide a basis for adjusting the instruction set. Through various sensors deployed in the fermentation chamber, such as temperature and humidity sensors and gas sensors, environmental data, including temperature, humidity, carbon dioxide concentration, and volatile metabolite concentration, is obtained every 15 minutes, and the changing trends of the obtained data are observed. Thus, the latest information in the fermentation process can be promptly obtained, providing data support for precise control, improving the monitoring ability of the fermentation process. Accurate real-time data and analysis results provide a basis for determining whether and how to adjust the control instruction set.

[0135] Clarify the gap between the current fermentation state and the preset target, judge the regulation effect, and determine whether it is necessary to adjust the regulation instruction set to ensure that the fermentation process develops in the expected direction; compare the real-time acquired and analyzed data with the preset fermentation target, where the preset fermentation target includes the appropriate fermentation temperature range, humidity range, and the expected value of the production rate of the target product, etc., and calculate the deviation between the actual data and the value of the fermentation target; thus intuitively judge the execution effect of the regulation instruction set, timely discover the deviation in the fermentation process, provide a quantitative basis for adjusting the regulation strategy, and the evaluation result determines whether to initiate the adjustment process of the regulation instruction set. If there is a deviation, it guides the subsequent adjustment steps.

[0136] According to the comparison and evaluation results, determine the specific adjustment direction and amplitude to make the regulation instruction set more in line with the fermentation requirements and optimize the fermentation process; if the temperature is too high, determine the adjustment strategy according to the size of the deviation. If the temperature deviation is between +1 and +2 °C, the heat input can be reduced by first decreasing the opening degree of the ventilation valve by 5% and observing the temperature change; if the deviation is greater than +2 °C, then simultaneously reduce the heating power by 1 kW and increase the opening degree of the ventilation valve by 10%. The adjustment of parameters such as humidity and carbon dioxide concentration also follows a similar logic. According to the deviation range and change trend of different parameters, corresponding adjustment strategies are formulated, and these strategies are pre-stored in the strategy library, and the specific adjustment plan is determined by calling the data in the strategy library; thus, the regulation instructions are adjusted specifically to make the fermentation environment closer to the ideal state, improve the fermentation efficiency and product quality, and the determined adjustment strategy provides the specific content for generating a new regulation instruction set, which is the key step in adjusting the regulation instruction set.

[0137] Convert the adjustment strategy into specific regulation instructions to achieve precise control of the fermentation equipment and ensure that the fermentation process proceeds according to the optimized plan; according to the determined adjustment strategy, generate a new regulation instruction set. For example, if the adjustment strategy is to reduce the heating power by 1 kW, increase the opening degree of the ventilation valve by 10%, and increase the humidification amount by 0.5 L / h, convert these adjustments into control instructions recognizable by the equipment, and transmit the instructions through the network to the regulation system of the fermentation equipment. The regulation system adjusts the operating parameters of the heating equipment, ventilation equipment, and humidification equipment according to the instructions; thus, precise control of the fermentation equipment is achieved, the fermentation environment is effectively adjusted, and the stability and efficiency of the fermentation process are guaranteed. The new regulation instruction set is sent to the fermentation equipment for execution, and at the same time, it provides a new starting point for subsequent monitoring and re-evaluation.

[0138] Check whether the adjusted control instruction set achieves the expected effect, provide feedback for subsequent further optimization, and ensure the effectiveness and sustainability of the control; after the new control instruction set is executed for a period of time (for example, 30 minutes), obtain and analyze the environmental data again, compare the data before and after the adjustment, and evaluate whether the various parameters are close to the preset targets, such as observing whether the temperature is reduced to the target range after adjustment. If the expected effect is achieved, continue to monitor according to the current control instruction set. If the expected effect is not achieved, re-evaluate the deviation, adjust the strategy, and generate a new control instruction set again to form a closed-loop feedback adjustment mechanism; thereby continuously optimizing the control instruction set, improving the stability and controllability of the fermentation process, and ensuring the consistency of product quality. The effect verification results determine whether the control instruction set needs to be adjusted again, which affects the control direction and rhythm of the entire fermentation process.

[0139] Clarify the parameter range of normal fermentation, provide judgment criteria for timely detection of abnormal situations, and ensure that measures can be taken quickly when abnormalities occur to avoid serious impacts on the fermentation process and product quality; set reasonable abnormal thresholds for each monitoring parameter based on a large amount of experimental data and production experience, such as the upper limit of temperature is set to 35°C and the lower limit is set to 20°C; the upper limit of humidity is set to 80% and the lower limit is set to 50%; the upper limit of carbon dioxide concentration is set to 0.6% and the lower limit is set to 0.05%. Store these abnormal thresholds in the database as a basis for abnormal judgment; thereby providing quantitative standards for abnormal monitoring, improving the timeliness and accuracy of abnormal discovery, and ensuring the safety and stability of the fermentation process. The abnormal threshold is the basis of abnormal detection and determines whether the subsequent abnormal handling process is started.

[0140] Monitor various parameters in the fermentation process in real time, and promptly detect abnormal situations beyond the normal range, providing guarantees for rapid response and handling of abnormalities; in the process of data acquisition and transmission, compare the acquired environmental data with the preset abnormal threshold in real time. Once a parameter is found to be beyond the threshold range, the abnormal alarm mechanism is immediately triggered. For example, when the temperature sensor detects that the temperature reaches 36°C, a high temperature alarm signal is issued; thereby capturing abnormal situations in a timely manner, buying time for subsequent processing, reducing the harm of abnormalities to the fermentation process, and reducing production losses.

[0141] Accurately determine the cause of the abnormality and provide a basis for taking targeted solutions to ensure that the abnormality can be effectively eliminated and normal fermentation can be restored. When the abnormal alarm is triggered, automatically collect relevant data within a period of time before and after the abnormality occurs, including environmental data and equipment operation data, and analyze the collected data. If the temperature rises abnormally, it is determined that the temperature abnormality is caused by a failure of the heating equipment or poor ventilation. In this way, the cause of the abnormality can be quickly and accurately determined, the pertinence and efficiency of abnormal handling can be improved, the time for abnormal handling can be reduced, and the impact on production can be reduced.

[0142] According to the cause of the abnormality, take corresponding measures to eliminate the abnormality, resume the normal progress of the fermentation process, and ensure product quality and production safety; if the cause of the abnormality is that the heating equipment fails and the temperature is too high, immediately stop the operation of the heating equipment, start the standby cooling equipment to reduce the temperature, and at the same time send a fault notice to the maintenance personnel, informing the type and location of the fault for timely repair. If the abnormality is caused by poor ventilation, automatically increase the opening of the ventilation valve and check whether the ventilation duct is blocked. If there is a blockage, clean it in time; thus effectively solve the abnormality problem, restore the stability of the fermentation environment, reduce the impact of the abnormality on product quality, and ensure the continuity of production. The result of the abnormality handling affects whether the fermentation process can return to normal. If the handling is successful, continue with normal fermentation monitoring. If the handling fails, further adjust the handling measures or suspend the fermentation for a comprehensive inspection.

[0143] Record the detailed information and handling process of the abnormality occurrence to provide data support for subsequent analysis and improvement, avoid similar abnormalities from occurring again, and improve the reliability and stability of the fermentation process; after the abnormality handling is completed, record the information such as the time of abnormality occurrence, abnormal parameters, cause of abnormality, handling measures, and handling results in the cloud knowledge base in detail. Regularly summarize and analyze the abnormality records to find out potential risk points and weak links. For example, if it is found that a certain batch frequently has humidity abnormalities caused by humidity sensor failures, the maintenance and replacement cycle of the humidity sensor can be optimized; thus accumulating experience in abnormality handling, providing a basis for optimizing the fermentation process and equipment maintenance, and improving the stability and reliability of the entire fermentation process.

Claims

1. An intelligent monitoring method for the fermentation process of dandelion golden camellia based on the Internet of Things, characterized in that: include: Deploy a sensor array to obtain environmental data, spectral data, and visual data in the fermentation chamber, and self-calibrate the sensor array; Extract the features of environmental data, spectral data and visual data to obtain the temperature and humidity time series characteristics, gas release rate, spectral characteristics and mycelium morphology characteristics, generate the mycelium growth rate based on the temperature and humidity time series characteristics, spectral characteristics and mycelium morphology characteristics, determine the bacterial community metabolic intensity based on the gas release rate and mycelium morphology characteristics, and predict the generation rate of the target product based on the spectral characteristics; The generation logic of the hyphae growth rate includes: The temperature and humidity time series characteristics, spectral characteristics and hyphae morphological characteristics at different time points were integrated and time-stamped and aligned; Select characteristic points on the mycelium skeleton and record the temperature and humidity time series characteristics and spectral characteristics of each characteristic point; The displacement of the feature point between two adjacent frames of hyphae morphology images is calculated to determine the local growth rate of the feature point in the time period, and the changes in the temperature and humidity time series characteristics and spectral characteristics in the time period are analyzed to determine the weight of the feature point; The local growth rates of all feature points are weighted averaged to obtain the hyphae growth rate; Generate fermentation stage labels and confidence levels based on mycelium growth rate and bacterial metabolic intensity, and generate control instruction sets based on reinforcement learning in combination with the target product production rate; Execute the control instruction set, monitor the environmental data to adjust the control instruction set and abnormal protection, and synchronize it to the cloud knowledge base.

2. The intelligent monitoring method for the fermentation process of Camellia chrysantha based on Internet of Things as claimed in claim 1, characterized in that: The hyphae morphological characteristics include hyphae density and hyphae coverage area; The extraction logic of mycelium morphological features includes: The hyphae morphology image is preprocessed, and the hyphae region in the hyphae morphology image is segmented according to the grayscale threshold; The segmented hyphae regions are skeletonized by a thinning algorithm to obtain a hyphae skeleton; Extract the topological structure of hyphae skeleton, including the number of branch points, the number of endpoints and the average length of hyphae branches; The hyphae density and hyphae coverage area were determined according to the topological structure of the hyphae skeleton.

3. The intelligent monitoring method for the fermentation process of Camellia chrysantha based on Internet of Things as claimed in claim 2, characterized in that: The logic for determining the bacterial flora metabolic intensity includes: Monitor the dynamic changes of carbon dioxide concentration and volatile metabolite concentration, and analyze the gas release rate; The gas release rate and hyphae morphological characteristics were correlated and analyzed; The metabolic intensity of the bacterial community was obtained by weighted quantification based on the release rate of carbon dioxide, the release rate of volatile metabolites and the change rate of mycelium coverage area.

4. The intelligent monitoring method for the fermentation process of Camellia chrysantha based on Internet of Things as claimed in claim 3, characterized in that: The prediction logic of the generation rate of the target product includes: Screening the bands of spectral characteristics related to the target product; The area of ​​the band is mapped to the substrate consumption rate, and the production rate of the output target product is predicted through a neural network in combination with environmental data; Monitor the prediction error of the neural network to determine whether to retrain the neural network.

5. The intelligent monitoring method for the fermentation process of Camellia chrysantha based on Internet of Things as claimed in claim 4, characterized in that: The generation logic of the fermentation stage label and confidence includes: The mycelium growth rate, bacterial community metabolic intensity and historical stage labels are integrated to obtain integrated data, and cluster analysis is performed on the integrated data based on a clustering algorithm to obtain clustering results; Based on fuzzy logic, the mycelium growth rate and bacterial metabolic intensity were divided into fuzzy linguistic variables, and the fermentation stage labels of different fuzzy linguistic variable combinations were determined based on the clustering results. The confidence of each fermentation stage label is calculated, and the fermentation stage label is corrected according to the confidence.

6. The intelligent monitoring method for the fermentation process of Camellia chrysantha based on Internet of Things as claimed in claim 5, characterized in that: The generation logic of the control instruction set includes: Construct a hierarchical reinforcement learning architecture, which includes a target determination phase and a regulation generation phase; Determine the reward function based on the target product generation rate, energy consumption and pollution risk; Monitor the operating data during the fermentation process and dynamically adjust the control generation stage and reward function.

7. The intelligent monitoring method for the fermentation process of Camellia chrysantha based on Internet of Things as claimed in claim 6, characterized in that: The target determination stage is used to determine the regulation target according to the fermentation stage label, confidence and the generation rate of the target product; The control generation phase is used to determine the state space and action space; The state space includes temperature, humidity, carbon dioxide concentration, mycelium density, bacterial metabolic intensity, and the production rate of target products; The action space includes heating power, ventilation valve opening and humidification amount.

8. The intelligent monitoring method for the fermentation process of Camellia chrysantha based on Internet of Things as claimed in claim 7, characterized in that: The environmental data include temperature, humidity, carbon dioxide concentration and volatile metabolite concentration, the spectral data are used to analyze the bacterial community metabolic intensity and the production rate of the target product, and the visual data include hyphae morphology images to capture the hyphae growth morphology; The deployment logic of the sensor array includes: Divide the fermentation tank into regional blocks according to its structural functions; In each area block, a sensor array is deployed in a layered and three-dimensional manner. The sensor array includes temperature and humidity sensors, gas sensors, spectral sensors, and visual sensors. The sensor array is synchronously stimulated to obtain environmental data, spectral data and visual data in the fermentation chamber respectively.

9. The method for intelligent monitoring of the fermentation process of Camellia chrysantha based on Internet of Things as claimed in claim 8, characterized in that: The self-calibration logic of the sensor array includes: Determine self-calibration conditions according to changes in time and data, and configure standard data; During the self-calibration process, the parameters of the sensor array are adjusted according to the difference between the acquired data and the standard data; After adjustment, verify the effect of self-calibration to provide feedback on whether to re-calibrate.

Images