A honeycomb type microorganism culture and mixing control method and device

Abstract

The application relates to the technical field of microbial culture, and discloses a honeycomb type microbial culture and mixing control method and device, which comprises the following steps: setting a distributed temperature sensor and a pressure sensor, collecting microbial colony image data, determining the growth size and maturity degree of the microorganisms, analyzing the growth size and maturity degree of the microorganisms, collecting the temperature of the microbial culture, judging whether the temperature of the microbial culture is abnormal, synchronously collecting the circulating water temperature in a preset period when it is judged that the temperature of the microbial culture is abnormal, judging whether the circulating water temperature for microbial culture incubation is abnormal, and adjusting the circulating water temperature when it is judged that the circulating water temperature is abnormal. The application improves the fine control requirement of the microorganisms and makes the control of the microbial culture more intelligent.

Description

Technical Field

[0001] This invention relates to the field of microbial culture technology, and more specifically, to a method and apparatus for honeycomb microbial culture and mixing control. Background Technology

[0002] Microbial culture, as a crucial foundational step in fields such as bioengineering, biofermentation, environmental remediation, and pharmaceutical manufacturing, places increasingly higher demands on the design and operational efficiency of culture devices. Conventional culture devices typically employ a planar layout with sparsely distributed culture containers, making it difficult to achieve modular, high-density structural arrangements and hindering large-scale integrated operations. Currently, most culture equipment has relatively rudimentary control systems, with limited precision in monitoring and controlling parameters such as temperature, pH, dissolved oxygen, and stirring speed, failing to achieve refined and dynamic process control. Multiple-batch operations and open system designs often increase the risk of microbial contamination, affecting the reproducibility of experimental results and the stability of industrial production.

[0003] Temperature is a key factor influencing microbial growth, metabolism, enzyme activity, and product formation. Traditional microbial culture devices mostly employ constant-temperature water baths, heating mantles, or incubators for temperature control. While these methods can achieve heating or temperature control of the culture environment, they suffer from problems such as delayed temperature response, uneven temperature distribution, and high energy consumption, making it difficult to meet the demands for high-precision, zoned control. Especially in multi-unit microbial culture systems with honeycomb structures, where the culture units are physically isolated, traditional overall temperature control methods struggle to account for temperature differences between units, potentially leading to localized overheating or underheating, thus affecting the stability and reproducibility of experimental results. Furthermore, with the increasing demand for refined microbial control, higher requirements are being placed on the real-time performance, intelligence, and local adjustment capabilities of temperature control systems.

[0004] Therefore, it is necessary to design a honeycomb microbial culture and mixing control method and device to solve the problems existing in the current technology. Summary of the Invention

[0005] In view of this, the present invention proposes a honeycomb microbial culture and mixing control method and device, which aims to solve the problems of sparse distribution of current culture containers, which are not conducive to large-scale integrated operation, uneven temperature distribution, and lag in temperature control response.

[0006] In one aspect, the present invention proposes a method for honeycomb microbial culture and mixing control, comprising:

[0007] Install distributed temperature and pressure sensors;

[0008] Microbial community image data is collected, and Canny edge detection is used to extract the contour information of multiple microorganisms in the image to determine the growth size and maturity of the microorganisms. Based on the convolutional neural network model, the image features are analyzed to determine the current growth size and maturity of the microorganisms, as well as the suitable temperature for microbial culture.

[0009] Based on the real-time temperature collected at various locations of the microbial community by the distributed temperature sensor, it is determined whether the temperature of the microbial culture is abnormal.

[0010] When it is determined that the temperature of the microbial culture is abnormal, the circulating water temperature is collected within a preset time period and preprocessed. The preprocessed circulating water temperature is divided into several sliding time windows. Statistical features are extracted for each time window to form a temperature feature vector. Based on the temperature feature vector, it is determined whether the circulating water temperature is abnormal.

[0011] The circulating water temperature includes the circulating inlet water temperature and the circulating outlet water temperature. When the circulating inlet water temperature is determined to be abnormal, the circulating inlet water temperature is adjusted.

[0012] Furthermore, image features are analyzed based on convolutional neural network models, including:

[0013] A historical information dataset was created by combining historical microbial profile information with their corresponding culture temperature thresholds.

[0014] The historical information dataset is sampled at a ratio of 4:1 to obtain a training subset and a test subset;

[0015] The neural network model is iteratively trained based on the training subset, the iteratively trained neural network model is evaluated based on the test subset, and it is determined whether to stop iterative training based on the evaluation value.

[0016] If the evaluation value of the neural network model after the current iteration is less than the evaluation value of the neural network model after the previous iteration, the magnitude of the change in the gradient direction of the neural network model is reduced, and iterative training continues until the preset number of iterations is reached; if the evaluation value of the neural network model after the current iteration is greater than or equal to the evaluation value of the neural network model after the previous iteration, iterative training is stopped, and the convolutional neural network model is obtained.

[0017] Furthermore, when using distributed temperature sensors to collect real-time temperatures at various locations within the microbial community to determine whether the temperature during microbial culture is abnormal, this includes:

[0018] A two-dimensional temperature map is established by collecting real-time temperature data from various locations of the microbial community using distributed temperature sensors. An arbitrary temperature data point is selected as the origin in the two-dimensional temperature map, and the correction range is determined by using the origin as the center and the correction radius.

[0019] The correction temperature is calculated based on the temperature data of each coordinate point within the correction range. When the difference between the temperature data of the origin and the correction temperature is greater than the temperature difference threshold, the origin is determined to be an abnormal data point, and the coordinates of the abnormal data point are recorded.

[0020] When the abnormal data point exists in the two-dimensional temperature graph, it is determined that the temperature data is abnormal;

[0021] When there are no abnormal data points in the two-dimensional temperature graph, it is determined that the temperature data is not abnormal.

[0022] Furthermore, the collected temperature data is preprocessed, and the preprocessed temperature data is divided into several sliding time windows; statistical features are extracted for each time window to form a temperature feature vector, including:

[0023] The preprocessing includes missing value imputation, normalization, and error value removal; the statistical features include maximum value, minimum value, mean, and standard deviation.

[0024] Furthermore, when determining whether the temperature of the circulating water used for microbial culture insulation is abnormal based on temperature feature vectors, it also includes:

[0025] Preset circulating water inlet temperature threshold and circulating water outlet temperature threshold; based on the preset circulating water inlet temperature threshold and circulating water outlet temperature threshold; determine whether the circulating water inlet temperature and circulating water outlet temperature are abnormal respectively;

[0026] When the circulating inlet water temperature is determined to be normal, but the circulating outlet water temperature is abnormal, an early warning is issued, and the warning type and warning level are determined.

[0027] When it is determined that there are no abnormalities in the circulating water inlet temperature and circulating water outlet temperature at the circulating water interface, an early warning is issued, and the warning type and warning level are determined.

[0028] Furthermore, when adjusting the circulating water temperature, the following should be considered:

[0029] Temperature feature vectors are extracted to establish a feature set. The feature set is then integrated with the historical feature set in the historical adjustment set to form a clustering set. The historical adjustment set includes several historical feature sets and several historical adjustment coefficients, and each historical feature set corresponds to a historical adjustment coefficient.

[0030] All sets to be clustered are processed using K-Means clustering, and the circulating water temperature is adjusted based on the clustering results.

[0031] Furthermore, when processing all sets to be clustered based on K-Means clustering and adjusting the circulating water temperature according to the clustering results, the process includes:

[0032] Normalize each data point in the set to be clustered;

[0033] S1: Initialize K centroids in the set to be clustered, and assign the remaining feature sets to the nearest centroids to form K clusters;

[0034] S2: Recalculate the centroid of each cluster;

[0035] S3: Repeat S1 and S2 until the centroid no longer changes;

[0036] S4: When the cluster containing the feature set does not contain a historical feature set, the historical adjustment coefficient corresponding to the maximum similarity between the feature set and the historical feature set is selected as the initial adjustment coefficient;

[0037] When the cluster containing the feature set includes a historical feature set, the average value of the historical adjustment coefficients corresponding to all the included historical feature sets is selected as the adjustment coefficient, and the circulating water temperature is adjusted according to the adjustment coefficient.

[0038] Furthermore, when selecting the historical adjustment coefficient corresponding to the maximum similarity between the feature set and the historical feature set as the initial adjustment coefficient, the following is also included:

[0039] The average circulating water temperature, the maximum circulating water temperature, and the minimum circulating water temperature are obtained based on the feature set, and the circulating water temperature difference is determined. The circulating water temperature difference is the difference between the maximum circulating water temperature and the average circulating water temperature, or the difference between the minimum circulating water temperature and the average circulating water temperature.

[0040] When the temperature difference of the circulating water corresponding to the largest absolute value is positive, a first correction coefficient is determined to correct the initial adjustment coefficient, and the circulating water temperature is adjusted with the corrected adjustment coefficient and cultured with the adjusted circulating water temperature. The first correction coefficient is inversely proportional to the temperature difference of the circulating water with the largest absolute value, and the value range of the first correction coefficient is (0.8, 1).

[0041] When the absolute value of the circulating water temperature difference corresponding to the largest value is negative, a second correction coefficient is determined to correct the initial adjustment coefficient, and the circulating water temperature is adjusted with the corrected adjustment coefficient and cultured with the adjusted circulating water temperature. The second correction coefficient is inversely proportional to the circulating water temperature difference with the largest absolute value, and the value range of the second correction coefficient is (1, 1.2).

[0042] Furthermore, it also includes:

[0043] The pressure value of the gas produced during the growth process of the microorganism is collected, and it is determined whether the pressure value exceeds the pressure threshold. When the pressure value exceeds the pressure threshold, the gas is released to the outside of the microbial growth environment to relieve pressure and maintain normal microbial culture.

[0044] Once all the microorganisms have been cultured, they are mixed together. After all the microorganisms have been mixed by rotating and stirring, the microbial mixing is complete.

[0045] Compared with existing technologies, the advantages of this invention are as follows: By comprehensively collecting temperature data from various regions of the microbial community through distributed temperature sensors and combining this with intelligent image recognition and analysis, a deep match between culture temperature and microbial growth status is achieved, ensuring that microorganisms grow in the most suitable temperature environment and improving culture efficiency and quality. Employing the Canny edge detection algorithm and convolutional neural network model, the invention can accurately identify the contours, size, and maturity of microorganisms. It dynamically determines the suitable culture temperature through a data-driven approach, and can quickly pinpoint whether an anomaly is caused by the circulating water system, improving the efficiency and accuracy of anomaly diagnosis. By using a sliding time window to divide the circulating water temperature data and extracting statistical features to construct a temperature feature vector, the invention can effectively capture temperature change trends and abnormal fluctuation patterns, improving the ability to judge circulating water anomalies and reducing false positives and false negatives. When the circulating water system is determined to be normal, an early warning is issued, and the warning type and level are determined by combining multi-source data, helping users to respond and handle the situation quickly, avoiding microbial culture failure or resource waste, and improving system stability and security. The invention integrates distributed sensors, image processing, and deep learning models to achieve full automation from data acquisition, status recognition, anomaly detection, and early warning response, reducing manual intervention.

[0046] On the other hand, this application also provides a honeycomb microbial culture and mixing device for applying the above-mentioned honeycomb microbial culture and mixing control method, including:

[0047] A culture tube assembly includes a culture tube body for microbial culture. The culture tube body has a pH sensor interface, a dissolved oxygen sensor interface, a temperature sensor interface, and a pressure sensor interface sequentially arranged on its side. The culture tube body has a water interface, a pressure control interface, a steam interface, and a circulating water interface arranged on its top. A quick-opening pressure cap is also provided on the top of the culture tube body. A one-way pressure valve is provided at the bottom of the culture tube body.

[0048] A mixing device includes a mixing tank, a stirring device, a motor, and a discharge port; the motor drives the stirring device to rotate and stir; the stirring device is used to stir and mix all microorganisms entering the mixing tank; the mixing tank is provided with at least one culture tube assembly;

[0049] A water supply device includes a water pump and a water tank; the water tank is provided with an inlet and an outlet at its upper and lower ends, respectively, and the water tank is used to store water for use; the outlet is connected to the water usage interface.

[0050] The monitoring device includes a pH sensor, a dissolved oxygen sensor, a temperature sensor, and a pressure sensor. Each sensor is connected to the corresponding interface of the pH sensor, dissolved oxygen sensor, temperature sensor, and pressure sensor, respectively, for monitoring and collecting various data within the culture tube body.

[0051] The control device is electrically connected to the water interface, air pressure control interface, steam interface and circulating water interface, and is used to control the opening degree of each interface; it is also electrically connected to the monitoring device and is used to receive the electrical signals from the monitoring device.

[0052] Compared with existing technologies, the advantages of this invention are as follows: By sequentially arranging pH sensor, dissolved oxygen sensor, temperature sensor, and pressure sensor interfaces on the side of the culture tube body, and in conjunction with the corresponding sensors, real-time and accurate monitoring of key parameters in the culture environment can be achieved, ensuring the stability and controllability of the microbial culture process. The system integrates a water interface, a pressure control interface, a steam interface, and a circulating water interface on the top of the culture tube body, allowing the system to flexibly connect to various auxiliary media to meet the needs of different culture conditions, improving culture efficiency and flexibility. The safe and reliable design of the quick-opening pressure cap and one-way pressure valve not only facilitates quick opening and closing by operators, ensuring operational safety, but also effectively prevents accidents caused by abnormal pressure, improving system safety. The motor-driven stirring device rotates, achieving thorough stirring of microorganisms in the mixing tank, avoiding sedimentation and uneven local environments, ensuring sufficient contact between the culture medium and microorganisms, and promoting uniform microbial growth. The water supply device is equipped with a water tank and a water pump to ensure a continuous supply of water, and the outlet is directly connected to the water interface of the culture tube, achieving automated and continuous water supply and simplifying the operation process. The control device is electrically connected to multiple interfaces and monitoring devices, and can automatically adjust the supply of air pressure, steam, circulating water and used water according to the data collected by the sensors, so as to realize the automatic adjustment and optimization of the culture environment, and improve the intelligence level and ease of operation of the culture process. Attached Figure Description

[0053] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:

[0054] Figure 1 This is a control flowchart of the honeycomb microbial culture and mixing control method provided in an embodiment of the present invention;

[0055] Figure 2 This is a structural block diagram of a honeycomb microbial culture and mixing device provided in an embodiment of the present invention;

[0056] Figure 3 This is a schematic diagram of a single culture tube assembly of a honeycomb microbial culture and mixing device provided in an embodiment of the present invention;

[0057] Figure 4 This is a top view of multiple culture tube components in a honeycomb microbial culture and mixing device provided in an embodiment of the present invention.

[0058] Figure 5 This is a schematic diagram of the culture tube assembly structure provided in an embodiment of the present invention.

[0059] The components include: 1. Culture tube assembly; 101. Culture tube body; 102. pH sensor interface; 103. Dissolved oxygen sensor interface; 104. Temperature sensor interface; 105. Pressure sensor interface; 106. Water interface; 107. Gas pressure control interface; 108. Steam interface; 109. Circulating water interface; 110. Quick-opening pressure cap; 111. One-way pressure valve; 2. Mixing device; 201. Mixing tank; 202. Stirring device; 203. Motor; 204. Discharge interface; 3. Water supply device; 301. Water pump; 302. Water tank; 303. Inlet; 304. Outlet. Detailed Implementation

[0060] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the disclosure to those skilled in the art. It should be noted that, unless otherwise specified, embodiments and features in the embodiments of the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0061] With the development of synthetic biology, multi-species co-culture technology, and the simulation of complex micro-ecosystems, higher demands are placed on the precise control, high-throughput operation, heterologous co-culture, and environmental regulation of microbial culture processes. Traditional microbial culture systems mainly consist of constant-temperature shakers, fermenters, or solid-state culture devices. While they offer good stability and reproducibility in culturing single microorganisms, they have significant shortcomings in the following situations: Multi-species co-culture is difficult; different microorganisms often have different optimal growth temperatures, pH levels, nutrient requirements, and metabolic characteristics, easily leading to competition, antagonism, or metabolic inhibition in the same culture environment, severely affecting co-culture results. Temperature control response is slow and lacks precision. Traditional temperature control systems (such as water baths and jacketed heaters) have slow response and high inertia during temperature adjustment, making it difficult to achieve rapid temperature rise and fall in localized areas, especially unsuitable for temperature-sensitive synthetic biology experiments. System integration is low; existing equipment is mostly a combination of single-function components, requiring separate settings for culture, temperature control, and mixing, resulting in cumbersome operation, poor scalability, and unsuitability for high-throughput experiments or automated control.

[0062] This shows that traditional microbial culture and mixing methods have significant limitations in multi-strain co-culture and precise environmental control.

[0063] In view of the above issues, please refer to Figure 1 As shown, this application proposes a method for honeycomb microbial culture and mixing control, including:

[0064] S100: Equipped with distributed temperature and pressure sensors;

[0065] S200: Collect microbial community image data, use Canny edge detection to extract the contour information of multiple microorganisms in the image, and determine the growth size and maturity of the microorganisms; analyze the image features based on the convolutional neural network model to determine the current growth size and maturity of the microorganisms, as well as the suitable temperature for current microbial culture;

[0066] S300: Based on the real-time temperature of the microbial community collected by the distributed temperature sensor, determine whether the temperature of the microbial culture is abnormal;

[0067] S400: When it is determined that the temperature of the microbial culture is abnormal, the circulating water temperature is collected within a preset time period, and the circulating water temperature is preprocessed. The preprocessed circulating water temperature is divided into several sliding time windows. Statistical features are extracted for each time window to form a temperature feature vector. Based on the temperature feature vector, it is determined whether the circulating water temperature is abnormal.

[0068] S500: The circulating water temperature includes the circulating inlet water temperature and the circulating outlet water temperature. When the circulating inlet water temperature is determined to be abnormal, the circulating inlet water temperature is adjusted.

[0069] Specifically, in the S100, the distributed temperature sensor uses a distributed fiber optic temperature sensor, which can measure temperature distribution along the length of the fiber, suitable for long-distance, multi-point monitoring. Its use of fiber optic sensing technology enables online distributed measurement of more parameters. Using a distributed fiber optic temperature sensor to detect the temperature of the microbial community, a single fiber can achieve continuous temperature distribution measurement, unlike traditional point sensors (such as thermocouples) which can only measure at fixed locations. It can achieve data acquisition at a resolution of meters or even higher, covering the entire culture tank or culture channel. It can detect the temperature gradient distribution in the microbial culture system, promptly identifying local overheating or uneven cooling, which helps prevent microbial imbalance. The pressure sensor, preferably a piezoresistive pressure sensor, is common and sensitive, suitable for detecting minute pressure changes. The sensor is installed in the sealed cavity of the culture container, ensuring real-time reflection of pressure changes in the culture environment. Digital signals can be read by using it in conjunction with the data acquisition system. The pressure changes of gases or liquids produced during microbial metabolism are often small. Sensors respond rapidly to minute pressure fluctuations, exhibiting high sensitivity and the ability to reflect pressure changes in real time. They enable continuous real-time monitoring without disrupting the culture system or opening the culture container. Their rapid response to pressure changes makes them suitable for monitoring and analyzing dynamic microbial reaction processes (such as gas production and metabolic peaks). Microbial gas production causes a pressure increase, which sensors can detect, thus reflecting fermentation activity. Even in closed, small culture flasks, pressure changes during metabolism can be monitored, indirectly assessing microbial growth.

[0070] Specifically, in S200, the collected microbial community image data undergoes image preprocessing (grayscale conversion, noise reduction, and contrast enhancement), Canny edge detection, contour extraction, contour filtering and microbial segmentation, feature extraction (size, area, roundness), maturity determination (based on morphological or color features), and statistical analysis to form visualized data, thereby determining the size and maturity of microorganisms; for example, yeast cells have a length of 6 micrometers, a roundness of 0.85, and an area of ​​28.27 µm. 2 The yeast was determined to be in the logarithmic growth phase. Visual data representing the size and maturity of the microorganism were then fed into a convolutional neural network model. Based on the trained model, the optimal temperature for microbial culture was calculated. For example, the optimal temperature for the growth of the yeast in the logarithmic growth phase is 25°C to 30°C.

[0071] Specifically, in the S300, distributed fiber optic temperature sensors are laid out in a staggered pattern on the surface, bottom, and sidewalls of the microbial culture area or container. A "honeycomb" or "grid" wiring method ensures that all local locations are monitored. Backscattered signals are collected in real time by emitting pulsed light from a laser that propagates along the fiber optic cable. The spatial location and temperature value of each point are calculated by combining this with the optical path time. For example, the temperature control accuracy is ±0.1℃, and the spatial location is set to 0.01 square meters. The system sets a sampling period, and each fiber optic sensing unit forms a continuous temperature profile. Through system calibration, a three-dimensional data mapping of temperature, location, and time is achieved. For example, the output format is T(x, t), representing the temperature value T at location x at time t. The detected real-time temperature is compared with the maximum and minimum temperature thresholds required for the microorganism's maturity and growth at that moment to confirm whether the temperature is abnormal. The maximum and minimum values ​​of the temperature range required for the microorganism's maturity and growth at that moment are the maximum and minimum temperature thresholds, respectively. When the real-time temperature is lower than the minimum temperature threshold or higher than the maximum temperature threshold, it is confirmed as a low-temperature anomaly or a high-temperature anomaly, respectively. When the real-time temperature is between the minimum and maximum temperature thresholds, there is no abnormality in the real-time temperature. For example, the suitable temperature for yeast growth in the logarithmic growth phase is 25℃ to 30℃. At this temperature, the minimum and maximum temperature thresholds for yeast are 25℃ and 30℃, respectively. When the detected temperature is 23℃, it is confirmed as a low-temperature abnormality; when the detected temperature is 35℃, it is confirmed as a high-temperature abnormality; and when the detected temperature is 26℃, the real-time temperature is confirmed to be normal.

[0072] Specifically, the S400 system monitors the microbial culture temperature in real time. When the real-time culture temperature exceeds the preset normal range (real-time temperature below the minimum threshold or above the maximum threshold), a temperature anomaly is determined. This triggers the synchronous collection of circulating water temperature data within a preset time period. The preset time and sampling frequency are set according to actual conditions. For example, the preset time can be 5 minutes or 30 minutes; the sampling frequency can be 1 second / time or 1 minute / time. The temperature data within the preset time period is preprocessed to remove obvious anomalies or fill in missing values. The preprocessed temperature data is divided into sliding time windows, and the length of each individual time window is determined (e.g., 1 minute or 3 minutes). The window length must balance timeliness and data representativeness. Window step size is set; windows can overlap, and the sliding step size can be smaller than the window length (e.g., a 3-minute window with a 30-second step size) to achieve smooth and continuous monitoring. The preprocessed time series data is cut according to the set window and step size to generate multiple continuous time window data segments. Statistical feature data is collected within these multiple time window data segments to form a temperature feature vector. The circulating water temperature is then compared with each threshold within the feature vector to determine if it is abnormal. When any value within the temperature feature vector exceeds the threshold range, the circulating water temperature is determined to be abnormal. When no value within the temperature feature vector exceeds the threshold, the circulating water temperature is determined to be normal. Thresholds are set for each feature based on experience or historical data. For example, the temperature feature vector includes the mean and standard deviation; the mean threshold range is 20℃-25℃, and the standard deviation threshold range is ≤12℃. When the mean is 26℃ and the standard deviation is 10℃, the mean exceeds the mean threshold range, and the circulating water temperature is determined to be abnormal. When the mean is 24℃ and the standard deviation is 10℃, neither the mean nor the standard deviation exceeds the corresponding threshold, and the circulating water temperature is determined to be normal.

[0073] Specifically, in S500, when the circulating water temperature is determined to be abnormal, it means that the circulating water temperature is already abnormal and there is no need to judge the circulating water temperature. After the circulating water temperature is abnormal, the circulating water temperature needs to be adjusted until the microbial culture temperature reaches the normal value.

[0074] Understandably, by combining image edge detection and convolutional neural networks, the system can identify the growth size and maturity of microorganisms in real time, eliminating the need for traditional manual microscopic observation and significantly improving efficiency and accuracy. Dynamically determining the optimal culture temperature range for microorganisms through image recognition results makes the temperature control strategy more targeted, overcoming the limitations of traditional temperature control systems with their fixed setpoints. Using distributed temperature sensors to acquire temperatures in multiple key areas within the culture tank in real time effectively avoids the risk of distortion from single-point measurements, providing a more accurate reflection of the temperature field distribution. Dividing the circulating water temperature data into sliding time windows and extracting statistical features (such as mean and standard deviation) effectively captures subtle temperature fluctuations and potential anomalies, improving the response sensitivity of the temperature control system. The system can distinguish between abnormal culture environment conditions and circulating water system malfunctions, avoiding blindly adjusting all temperature control modules and helping to quickly pinpoint the source of the anomaly, reducing false alarms or unnecessary repairs. By prioritizing the determination of abnormal circulating water temperature, core problems in the temperature control system can be identified promptly, avoiding resource waste and response delays caused by repeatedly judging the outlet water temperature. When the influent water temperature is abnormal, directly adjusting this parameter to the normal temperature suitable for microbial culture helps to quickly restore system stability, ensuring the consistency and effectiveness of the culture environment, thereby improving overall temperature control efficiency and biological culture quality. The image recognition and temperature control analysis algorithm can be trained and optimized according to different microbial species, applicable to various industrial and scientific research applications such as yeast, lactic acid bacteria, Escherichia coli, actinomycetes, and fungi. The modular architecture design allows for flexible integration with existing culture devices, data acquisition systems, and SCADA control platforms, providing excellent compatibility and scalability. Precise temperature control improves microbial growth rates, promptly detects temperature control anomalies, and allows for early intervention. Image analysis and automatic early warning reduce the frequency of manual inspections. Temperature stability directly affects the quality of biological products, and precise temperature control reduces energy consumption, aligning with energy conservation and emission reduction goals.

[0075] In some embodiments of this application, image feature analysis is performed based on a convolutional neural network model, including:

[0076] A historical information dataset was created by combining historical microbial profile information with their corresponding culture temperature thresholds.

[0077] The historical information dataset is sampled at a ratio of 4:1 to obtain a training subset and a test subset;

[0078] The neural network model is iteratively trained based on the training subset, the iteratively trained neural network model is evaluated based on the test subset, and it is determined whether to stop iterative training based on the evaluation value.

[0079] If the evaluation value of the neural network model after the current iteration is less than the evaluation value of the neural network model after the previous iteration, the magnitude of the change in the gradient direction of the neural network model is reduced, and iterative training continues until the preset number of iterations is reached; if the evaluation value of the neural network model after the current iteration is greater than or equal to the evaluation value of the neural network model after the previous iteration, iterative training is stopped, and the convolutional neural network model is obtained.

[0080] Understandably, by modeling the correspondence between historical microbial image contours and temperature thresholds, this invention can automatically identify the optimal culture temperature range for different types and growth stages of microorganisms. Compared to traditional methods relying on experience or manual measurement, this significantly improves the accuracy and intelligence of temperature control. This method introduces an evaluation feedback mechanism during training. By comparing the current model evaluation value with the previous evaluation value, it determines whether to reduce the learning rate and continue training or stop training early, thereby avoiding overfitting or oscillating decline in the later stages of training and effectively improving the model's generalization ability and stability. A 4:1 sampling method is used to divide the historical dataset into training and test sets, ensuring the breadth of training data and the objectivity of test data. This method saves training time and effectively evaluates the model's performance on unseen data, making the training process more scientific and efficient. A model performance monitoring mechanism is set up during training: when the test set evaluation value does not improve or even decreases, the training termination mechanism is automatically triggered. Compared to the traditional method of fixed iterations, this effectively prevents the model from overfitting the training set, thereby improving the model's adaptability and generalization ability to new microbial samples. When model performance falls short of expectations and evaluation values ​​decline, the gradient step size is automatically reduced (e.g., the learning rate is lowered), allowing for more precise and detailed model adjustments, avoiding getting trapped in local optima, and improving the model's final convergence. This mechanism is more intelligent and practical than traditional gradient descent algorithms. Leveraging the deep feature extraction capabilities of convolutional neural networks, key parameters such as the edges, morphology, density, and structure of microbial contours in images can be efficiently extracted, establishing a nonlinear mapping model from image to temperature, suitable for complex and diverse microbial culture environments.

[0081] In some embodiments of this application, when determining whether the temperature of the microbial culture is abnormal by collecting real-time temperatures at various locations within the microbial community using distributed temperature sensors, the method includes:

[0082] A two-dimensional temperature map is established by collecting real-time temperature data from various locations of the microbial community using distributed temperature sensors. An arbitrary temperature data point is selected as the origin in the two-dimensional temperature map, and the correction range is determined by using the origin as the center and the correction radius.

[0083] The correction temperature is calculated based on the temperature data of each coordinate point within the correction range. When the difference between the temperature data of the origin and the correction temperature is greater than the temperature difference threshold, the origin is determined to be an abnormal data point, and the coordinates of the abnormal data point are recorded.

[0084] When the abnormal data point exists in the two-dimensional temperature graph, it is determined that the temperature data is abnormal;

[0085] When there are no abnormal data points in the two-dimensional temperature graph, it is determined that the temperature data is not abnormal.

[0086] Specifically, multiple temperature sensor detection points are first deployed inside or on the surface of the microbial culture area. These detection points are evenly distributed on a two-dimensional plane and collect real-time temperature data according to their physical spatial coordinates (x, y). The temperature value collected by each detection point is mapped to its position in space to generate a complete two-dimensional temperature map, i.e., a two-dimensional matrix composed of temperature data. For example, a point T(x=3, y=5) = 37.2℃ in the temperature map indicates that the sensor at coordinates (3, 5) measures a temperature of 37.2℃. An arbitrary coordinate point is selected from the temperature map as the starting point for analysis, i.e., the "origin". Let the coordinates of this origin be (x1, y1), and its temperature value be T1. To determine whether the temperature at this point is reasonable, it is necessary to analyze the surrounding temperature data. A fixed correction radius R is set with the origin as the center, forming a circular area centered on the origin (i.e., the correction range). This area includes all points whose distance from the origin does not exceed R, constituting the set of neighboring points used for correction reference. Temperature data of all neighboring points are collected from this correction range (excluding the origin itself). Averaging these temperature values ​​yields the corrected temperature value Tc for the region, which can be expressed as: Corrected temperature Tc = Sum of temperatures of all neighboring points / Number of neighboring points. This temperature value represents the "normal" temperature level within this local area. The origin temperature T1 is compared with the corrected temperature Tc, and the temperature difference ΔT is calculated: ΔT = |T1 - Tc|. A predefined temperature difference threshold ΔTthresh is set. If ΔT exceeds this threshold (ΔT > ΔTthresh), it indicates a significant deviation between the temperature of this point and its neighboring points, potentially due to sensor malfunction or localized abnormal heating / cooling. In this case, the origin is determined to be an anomalous data point, and its spatial coordinates (x1, y1) are recorded for subsequent analysis.

[0087] Understandably, by accurately calculating the temperature difference between each coordinate point and its neighborhood, this method can keenly detect local overheating or overcooling phenomena, promptly identify non-uniform temperature control problems caused by equipment failure, uneven heat transfer, or environmental fluctuations, and detect abnormal trends in advance, providing early warning for microbial culture and helping to avoid culture failure or product abnormalities. Traditional temperature control methods often rely on single-point or average temperatures as the basis for judgment, which is difficult to reflect details. However, this method makes multi-point judgments based on local coordinate temperature differences, which is more sensitive than overall averaging. It supports the localization of thermal anomalies in two-dimensional space, improves the real-time adjustment capability of temperature control strategies, and can serve as the algorithm core of intelligent temperature control systems to achieve closed-loop control.

[0088] In some embodiments of this application, the collected temperature data is preprocessed, and the preprocessed temperature data is divided into several sliding time windows; statistical features are extracted for each time window to form a temperature feature vector, including:

[0089] The preprocessing includes missing value imputation, normalization, and error value removal; the statistical features include maximum value, minimum value, mean, and standard deviation.

[0090] Understandably, preprocessing techniques such as missing value imputation and error value removal effectively address the issues of null and erroneous values ​​in the original temperature data caused by sensor malfunctions and communication interruptions. This makes the data more complete and continuous, reduces data noise, and provides a stable foundation for subsequent analysis. Normalization unifies temperature data collected at different times and in different regions to the same numerical scale, eliminating data bias caused by differences in equipment, environment, or units, thereby ensuring comparability between data and facilitating unified model processing. By extracting statistical features such as maximum, minimum, mean, and standard deviation in each time window, the original one-dimensional temperature time series is transformed into a structured multi-dimensional feature vector, significantly enhancing the data's ability to express temperature change characteristics and providing richer information support for subsequent classification and clustering.

[0091] In some embodiments of this application, when determining whether the temperature of the circulating water used for microbial culture insulation is abnormal based on the temperature feature vector, the method further includes:

[0092] Preset circulating water inlet temperature threshold and circulating water outlet temperature threshold; based on the preset circulating water inlet temperature threshold and circulating water outlet temperature threshold; determine whether the circulating water inlet temperature and circulating water outlet temperature are abnormal respectively;

[0093] When the circulating inlet water temperature is determined to be normal, but the circulating outlet water temperature is abnormal, an early warning is issued, and the warning type and warning level are determined.

[0094] When it is determined that there are no abnormalities in the circulating water inlet temperature and circulating water outlet temperature at the circulating water interface, an early warning is issued, and the warning type and warning level are determined.

[0095] Specifically, both the preset circulating inlet water temperature threshold and the circulating outlet water temperature threshold include a minimum temperature threshold and a maximum temperature threshold. The specific values ​​of the minimum or maximum circulating inlet water temperature threshold are determined based on the heat exchange efficiency during microbial cultivation and the required cultivation temperature of the microorganisms at the current time (the ratio of the required cultivation temperature of the microorganisms at the current time to the heat exchange efficiency). For example, if the minimum and maximum cultivation temperatures for yeast are 25℃ and 30℃ respectively, and the heat exchange efficiency is 60%, the minimum circulating inlet water temperature threshold is 41.7℃ (25℃ / 0.6), and the maximum circulating inlet water temperature threshold is 50℃ (30℃ / 0.6). The specific values ​​of the minimum or maximum circulating outlet water temperature threshold are determined based on the minimum or maximum circulating inlet water temperature threshold and the heat exchange efficiency. For example, if the minimum or maximum circulating inlet water temperature thresholds are 40℃ and 50℃ respectively, and the heat exchange efficiency is 70%, the minimum circulating inlet water temperature threshold is 12℃ (40℃ / 1-0.7), and the maximum circulating inlet water temperature threshold is 15℃ (50℃ / 1-0.7). When the circulating inlet water temperature is determined to be normal, but the circulating outlet water temperature is abnormal, it indicates an anomaly in the heat exchange of the entire microbial community, triggering an early warning. The circulating outlet water temperature may fall below the low-temperature threshold or exceed the high-temperature threshold, thus classifying the warning as a heat exchange anomaly warning. The warning level is determined based on the actual temperature conditions, specifically the difference between the detected real-time microbial culture temperature and the microbial threshold, along with the assigned warning level. Conversely, if the circulating water temperature at the circulating water interface is normal, confirming that the system cannot detect the cause of the abnormal microbial culture temperature, a warning is randomly issued, classifying the warning as a system anomaly. The warning level is also determined based on the actual temperature conditions, specifically the difference between the detected real-time microbial culture temperature and the microbial threshold, along with the assigned warning level. For example, the warning levels are divided into levels in increments of 10°C. When the minimum and maximum temperature thresholds for microbial culture are 25°C and 30°C, respectively, the temperature range for the first low-temperature warning level is greater than or equal to 15°C and less than 25°C; the temperature range for the second low-temperature warning level is greater than or equal to 5°C and less than 15°C; the temperature range for the first high-temperature warning level is greater than 30°C and less than or equal to 40°C; and the temperature range for the second high-temperature warning level is greater than 40°C and less than or equal to 50°C.

[0096] Understandably, by constructing temperature feature vectors for the circulating inlet and outlet water, the system can accurately identify temperature change trends and abnormal states, avoiding misjudgments or omissions caused by relying solely on a single threshold in traditional methods. This enables more intelligent and detailed temperature control management. The method separately judges the abnormal states of the inlet and outlet water temperatures and executes corresponding operations based on different judgment results. For example, if the inlet water is abnormal, the temperature is automatically adjusted; if the outlet water is abnormal, a tiered warning is issued. This differentiated processing approach can more accurately pinpoint the source of the problem, avoid system misoperation, and improve response efficiency. When the system determines that the inlet water temperature is abnormal, it automatically executes the adjustment operation of the temperature control equipment, achieving rapid temperature correction without manual intervention, ensuring the stability of the microbial culture environment, and improving the continuity and self-repair capability of the overall system operation. When the outlet water temperature is abnormal while the inlet water is normal, the system issues a warning and classifies the warning type and level, helping to detect potential equipment failures, culture tank abnormalities, or abnormal microbial metabolism in advance, reducing potential risks and protecting the microbial growth environment.

[0097] In some embodiments of this application, adjusting the circulating water inlet temperature includes:

[0098] Temperature feature vectors are extracted to establish a feature set. The feature set is then integrated with the historical feature set in the historical adjustment set to form a clustering set. The historical adjustment set includes several historical feature sets and several historical adjustment coefficients, and each historical feature set corresponds to a historical adjustment coefficient.

[0099] All sets to be clustered are processed using K-Means clustering, and the circulating water temperature is adjusted based on the clustering results.

[0100] Specifically, this invention introduces a circulating water temperature adjustment mechanism based on the K-Means clustering algorithm. By combining historical data with the current operating status, it achieves intelligent and dynamic adjustment of the circulating water temperature. Compared to traditional methods that rely on fixed thresholds or empirical formulas for adjustment, this invention enables intelligent decision-making, improving the scientific rigor and accuracy of adjustments. By extracting the temperature feature vector of the current circulating water system and integrating it with historical adjustment data, the invention analyzes historical adjustment behaviors under similar operating conditions using the K-Means clustering algorithm. This effectively identifies the most similar historical scenarios to the current operating status, thereby achieving more targeted and adaptive temperature adjustment decisions. This avoids the problems of insufficient human experience or failure to handle complex operating conditions in traditional adjustment methods. Clustering can eliminate the interference of local noise and individual outliers on decision-making, making the adjustment strategy more robust. By statistically analyzing multiple historical adjustment coefficients from similar clusters (e.g., by taking a weighted average), the risk of misjudgment can be effectively reduced, improving the stability and accuracy of temperature control, and avoiding system temperature fluctuations or increased energy consumption due to over- or under-adjustment. As the system operates over time, it continuously accumulates historical feature sets and adjustment coefficients, forming a dynamically updated historical adjustment database. Combined with real-time data clustering, it can adapt to various complex changing conditions such as different seasons, loads, and external environments, significantly improving the versatility of the circulating water temperature control system. Precisely adjusting the circulating water temperature based on cluster analysis helps maintain microbial cultivation within its optimal thermal efficiency range, reducing unnecessary overheating or overcooling, thereby lowering heating or cooling energy consumption, improving the overall system's energy utilization rate, and demonstrating excellent energy-saving and consumption-reducing effects.

[0101] In some embodiments of this application, when processing all sets to be clustered based on K-Means clustering and adjusting the circulating water temperature according to the clustering results, the following steps are included:

[0102] Normalize each data point in the set to be clustered;

[0103] S1: Initialize K centroids in the set to be clustered, and assign the remaining feature sets to the nearest centroids to form K clusters;

[0104] S2: Recalculate the centroid of each cluster;

[0105] S3: Repeat S1 and S2 until the centroid no longer changes;

[0106] S4: When the cluster containing the feature set does not contain a historical feature set, the historical adjustment coefficient corresponding to the maximum similarity between the feature set and the historical feature set is selected as the initial adjustment coefficient;

[0107] When the cluster containing the feature set includes a historical feature set, the average value of the historical adjustment coefficients corresponding to all the included historical feature sets is selected as the adjustment coefficient, and the circulating water temperature is adjusted according to the adjustment coefficient.

[0108] Specifically, when adjusting the circulating water temperature, the current temperature characteristic data needs to be integrated with historical temperature characteristic data to form a dataset to be clustered. To ensure the accuracy and stability of the clustering analysis, each data point in the dataset is first normalized. This step aims to eliminate dimensional and numerical range differences between various temperature features, giving them a uniform numerical scale and avoiding clustering bias due to inconsistent data scales. After normalization, the K-Means clustering algorithm is used to analyze the normalized dataset. K points are randomly selected as initial centroids in the normalized dataset. Then, all remaining feature sets are assigned to the clusters with the nearest centroids based on their distance from each centroid, forming K initial clusters. The centroid position of each cluster is recalculated based on all feature sets within the newly assigned clusters. The centroid represents the center position of all data within a cluster and reflects the average characteristic state of that cluster. Repeat steps S1 and S2, continuously adjusting the affiliation of feature sets within clusters and recalculating centroids until the centroid positions stabilize and no longer change, indicating that the clustering process has converged. If the current feature set's cluster does not contain any historical feature sets, this indicates that the current operating state is unique or novel, with no direct historical data to correspond to it. In this case, the system calculates the similarity between the current feature set and all historical feature sets, selecting the historical adjustment coefficient corresponding to the historical feature set with the highest similarity as the current initial adjustment coefficient. If the current feature set's cluster contains one or more historical feature sets, this indicates that the current operating state has a strong similarity to these historical states. The system will statistically analyze the adjustment coefficients corresponding to all historical feature sets in the cluster and calculate their average value as the temperature adjustment coefficient for this adjustment. Finally, the system adjusts the circulating water temperature accordingly based on the obtained adjustment coefficients.

[0109] Understandably, by performing cluster analysis on current temperature characteristics and a large amount of historical characteristic data, intelligent temperature adjustment based on similar operating states is achieved, avoiding the limitations of traditional empirical methods or fixed threshold adjustments, and improving the scientific nature and accuracy of the adjustment. Clustering results are used to distinguish different operating conditions, enabling dynamic adjustment. Even when encountering new operating states, the method can calculate the adjustment coefficient of the most similar historical data to achieve reasonable initial adjustments, ensuring stable system operation. The entire adjustment process relies on data-driven cluster analysis, reducing operator reliance on experience and adjustment errors, effectively improving the system's automation and intelligence levels. By categorizing data with similar characteristics through clustering and averaging the adjustment coefficient within clusters, the impact of outliers is effectively weakened, preventing drastic fluctuations in temperature adjustment due to occasional anomalies and ensuring the stability of temperature control.

[0110] In some embodiments of this application, when selecting the historical adjustment coefficient corresponding to the maximum similarity between the feature set and the historical feature set as the initial adjustment coefficient, the method further includes:

[0111] The average circulating water temperature, the maximum circulating water temperature, and the minimum circulating water temperature are obtained based on the feature set, and the circulating water temperature difference is determined. The circulating water temperature difference is the difference between the maximum circulating water temperature and the average circulating water temperature, or the difference between the minimum circulating water temperature and the average circulating water temperature.

[0112] When the temperature difference of the circulating water corresponding to the largest absolute value is positive, a first correction coefficient is determined to correct the initial adjustment coefficient, and the circulating water temperature is adjusted with the corrected adjustment coefficient and cultured with the adjusted circulating water temperature. The first correction coefficient is inversely proportional to the temperature difference of the circulating water with the largest absolute value, and the value range of the first correction coefficient is (0.8, 1).

[0113] When the absolute value of the circulating water temperature difference corresponding to the largest value is negative, a second correction coefficient is determined to correct the initial adjustment coefficient, and the circulating water temperature is adjusted with the corrected adjustment coefficient and cultured with the adjusted circulating water temperature. The second correction coefficient is inversely proportional to the circulating water temperature difference with the largest absolute value, and the value range of the second correction coefficient is (1, 1.2).

[0114] Specifically, after selecting the historical adjustment coefficient corresponding to the historical feature set with the highest similarity to the current feature set as the initial adjustment coefficient, the system further refines the initial adjustment coefficient by analyzing the statistical characteristics of the current circulating water temperature, so as to improve the rationality and accuracy of temperature adjustment.

[0115] Specifically, the system first obtains the following from the current circulating water temperature data: the average circulating water temperature (reflecting the average level of circulating water temperature within the current time period), the maximum circulating water temperature (reflecting the upper limit of temperature fluctuation), and the minimum circulating water temperature (reflecting the lower limit of temperature fluctuation). Based on this data, the system calculates two temperature differences: the difference between the maximum and the average, and the difference between the minimum and the average. These two differences reflect the fluctuation range and direction of deviation from the average of the current circulating water temperature. The system then identifies the difference with the largest absolute value and executes different correction strategies based on the sign of this difference. When the largest absolute value difference is positive, it indicates that the maximum circulating water temperature is higher than the average, and there is a certain upward shift or peak in temperature. In this case, the system determines a parameter called the first correction coefficient to correct the initial adjustment coefficient. The first correction coefficient is inversely proportional to the maximum absolute difference; that is, the greater the temperature fluctuation, the closer the correction coefficient is to a lower value (but not lower than 0.8). When the fluctuation is small, the correction coefficient approaches 1. The value of this correction coefficient is limited to between 0.8 and 1. The system multiplies the initial adjustment coefficient by the first correction coefficient to obtain the corrected adjustment coefficient. The circulating water temperature is adjusted according to the corrected adjustment coefficient, and subsequent cultivation is carried out at this temperature. This ensures that when there is a large positive temperature fluctuation, the adjustment coefficient is appropriately reduced to avoid system oscillation or overshoot caused by excessive temperature adjustment. When the maximum absolute difference is negative, it indicates that the minimum circulating water temperature is lower than the average, and there is a certain downward shift or trough in the temperature. In this case, the system determines a parameter called the second correction coefficient to correct the initial adjustment coefficient. The second correction coefficient is also inversely proportional to the maximum absolute difference. The greater the temperature fluctuation, the closer the correction coefficient is to a higher value (but not exceeding 1.2); when the fluctuation is small, the correction coefficient approaches 1. The value range of this correction coefficient is limited to between 1 and 1.2. The system multiplies the initial adjustment coefficient by the second correction coefficient to obtain the corrected adjustment coefficient. The circulating water temperature is adjusted using this adjustment coefficient, and subsequent cultivation is carried out accordingly.

[0116] In some embodiments of this application, it also includes:

[0117] The pressure value of the gas produced during the growth process of the microorganism is collected, and it is determined whether the pressure value exceeds the pressure threshold. When the pressure value exceeds the pressure threshold, the gas is released to the outside of the microbial growth environment to relieve pressure and maintain normal microbial culture.

[0118] Once all the microorganisms have been cultured, they are mixed together. After all the microorganisms have been mixed by rotating and stirring, the microbial mixing is complete.

[0119] Specifically, during microbial culture, the system continuously monitors the pressure of gases produced during microbial growth and compares it to a preset pressure threshold. When the detected pressure exceeds the threshold, the system activates a pressure relief device to release some gas outside the microbial growth environment, thus automatically depressurizing and ensuring suitable pressure conditions within the culture environment to guarantee normal microbial growth and cultivation. After all microbial culture processes are completed, the system proceeds to the next stage of mixing. All cultured microorganisms are collected and introduced into a mixing unit. In this unit, different microorganisms are thoroughly mixed using a rotating agitator to ensure uniform distribution and achieve the predetermined mixing uniformity requirements. Once the system determines that the mixing process is complete, the mixing of all microorganisms is finished, and the process is complete, ready for subsequent processing or application.

[0120] Understandably, by monitoring the gas pressure generated during microbial growth in real time and promptly releasing pressure when it exceeds a set threshold, it is possible to effectively prevent damage to the microbial growth environment caused by excessive pressure, ensuring that the culture system is in a stable and suitable pressure state, and improving the success rate and consistency of microbial culture. By setting a pressure threshold and coordinating it with an automatic pressure release mechanism, intelligent control of gas pressure can be achieved without manual intervention, improving the automation level of system operation and reducing operational complexity and human error. After all microbial cultures are completed, efficient mixing is achieved through rotary stirring, ensuring uniform fusion of various microorganisms and avoiding the adverse effects of uneven mixing on subsequent applications or reaction processes, thereby improving overall culture quality and process controllability.

[0121] In another preferred embodiment based on the above embodiments, see [reference] Figure 2-5 As shown, the present invention also proposes a honeycomb microbial culture and mixing device for applying a honeycomb microbial culture and mixing control method, comprising:

[0122] The culture tube assembly 1 includes a culture tube body 101, which is used for microbial culture. A pH sensor interface 102, a dissolved oxygen sensor interface 103, a temperature sensor interface 104, and a pressure sensor interface 105 are sequentially arranged on the side of the culture tube body 101. A water interface 106, a gas pressure control interface 107, a steam interface 108, and a circulating water interface 109 are arranged on the top of the culture tube body 101. A quick-opening pressure cap 110 is also provided on the top of the culture tube body 101. A one-way pressure valve 111 is provided at the bottom of the culture tube body 101.

[0123] The mixing device 2 includes a mixing tank 201, a stirring device 202, a motor 203, and a discharge port 204; the motor 203 is used to drive the stirring device 202 to rotate and stir; the stirring device 202 is used to stir and mix all microorganisms entering the mixing tank 201; at least one culture tube assembly is provided on the mixing tank 201.

[0124] The water supply device 3 includes a water pump 301 and a water tank 302; the water tank 302 is provided with an inlet 303 and an outlet 304 at its upper and lower ends, respectively, and the water tank 302 is used to store water for use; the outlet 304 is connected to the water use interface 106.

[0125] The monitoring device includes a pH sensor, a dissolved oxygen sensor, a temperature sensor, and a pressure sensor. Each sensor is connected to the corresponding pH sensor interface 102, dissolved oxygen sensor interface 103, temperature sensor interface 104, and pressure sensor interface 105, respectively, for monitoring and collecting various data within the culture tube body 101.

[0126] The control device is electrically connected to the water interface 106, the air pressure control interface 107, the steam interface 108, and the circulating water interface 109, and is used to control the opening and closing degree of each interface; it is also electrically connected to the monitoring device and is used to receive the electrical signals from the monitoring device.

[0127] Specifically, the culture tube assembly 1 includes a culture tube body 101, the interior of which serves as the culture space for microorganisms. Four sensor interfaces are sequentially arranged on the side of the culture tube body 101: a pH sensor interface 102, a dissolved oxygen sensor interface 103, a temperature sensor interface 104, and a pressure sensor interface 105. Each interface is used to connect to the corresponding type of sensor for real-time monitoring of key parameters of the culture environment. When an abnormal value is detected, an alarm can be issued or automatic adjustments can be made via a control device, depending on the actual situation. Multiple functional interfaces are located on the top of the culture tube body, including a water interface 106 (for introducing water, including purified water or culture water), a pressure control interface 107 (for regulating internal pressure to maintain a stable culture environment), a steam interface 108 (for heating or sterilization), and a circulating water interface 109 (for temperature control during microbial culture). The top of the culture tube body 101 is equipped with a quick-opening pressure cap 110 for easy and rapid opening and sealing, ensuring operational safety and a sealed culture environment. A one-way pressure valve 111 is provided at the bottom. After cultivation, pressure is applied to the inside of the culture tube body 101 through the air pressure control interface 107, causing the microorganisms in each culture tube to enter the mixing device 2 below under pressure. The mixing device 2 includes a mixing tank 201, a stirring device 202, a motor 203, and a discharge interface 204. At least one culture tube assembly 1 is provided on the mixing tank 201. After the microorganisms in all culture tube assemblies 1 have completed cultivation and entered the mixing tank 201, the mixing tank 201 is used to hold the microbial culture solution to be mixed. The stirring device 202 is installed inside the mixing tank 201 and is driven to rotate by the motor 203. After stirring is completed, the uniformly mixed microbial culture solution is discharged through the discharge interface 204 for subsequent processing or transfer. The water supply device 3 includes a water tank 302 and a water pump 301. The water tank has an inlet 303 and an outlet 304 at its upper and lower ends, respectively, for the introduction and output of water. The water tank stores the required water (purified water or specific culture water). The outlet 304 is connected to the water interface 106 on the culture tube body 101. After the water pump 301 starts, it drives the water flow, ensuring a continuous and stable water supply within the culture tube to meet the water requirements of the cultivation process. The monitoring device includes various sensors, including a pH sensor, dissolved oxygen sensor, temperature sensor, and pressure sensor. These sensors are connected to the culture tube body through their corresponding sensor interfaces to monitor the environmental parameters inside the culture tube in real time. The data collected by the sensors is continuously collected and transmitted to ensure that key indicators such as pH, dissolved oxygen content, temperature, and pressure in the microbial culture environment are within the optimal range, supporting refined management of the cultivation process. The control device is electrically connected to the water interface 106, air pressure control interface 107, steam interface 108, and circulating water interface 109 on the culture tube body 101, and is also electrically connected to the monitoring device.The control device receives electrical signals of various environmental parameters from the monitoring device and automatically adjusts the opening and closing of each interface through built-in control logic to achieve automatic control of water flow, air pressure, steam supply, and circulating water flow, maintaining a stable and optimized culture environment. When the monitored parameters deviate from the preset threshold, the control device can automatically adjust the relevant interfaces or issue an alarm to ensure the safety and efficiency of the microbial culture process.

[0128] Understandably, by sequentially installing a pH sensor interface 102, a dissolved oxygen sensor interface 103, a temperature sensor interface 104, and a pressure sensor interface 105 on the side of the culture tube body 101, and in conjunction with the corresponding sensors, real-time and accurate monitoring of key parameters in the culture environment can be achieved, ensuring the stability and controllability of the microbial culture process. The culture tube body 101 is equipped with a water interface 106, a gas pressure control interface 107, a steam interface 108, and a circulating water interface 109, allowing for flexible connection to various auxiliary media to meet the needs of different culture conditions, improving culture efficiency and flexibility. The safe and reliable design of the quick-opening pressure cap 110 and the one-way pressure valve 111 not only facilitates quick opening and closing by operators, ensuring operational safety, but also effectively prevents accidents in case of abnormal pressure, improving system safety. The stirring device 202 is driven by a motor 203 to rotate, achieving thorough stirring of the microorganisms in the mixing tank 201, avoiding sedimentation and uneven local environments, ensuring sufficient contact between the culture medium and microorganisms, and promoting uniform microbial growth. The water supply device 3 is equipped with a water tank 302 and a water pump 301 to ensure a continuous supply of water. It is directly connected to the water interface 106 of the culture tube via the outlet 304, achieving automated and continuous water supply and simplifying the operation process. The control device is electrically connected to multiple interfaces and monitoring devices, and can automatically adjust the supply of air pressure, steam, circulating water, and water based on data collected by sensors. This enables automatic adjustment and optimization of the culture environment, improving the intelligence level and ease of operation of the culture process.

[0129] In summary, this application integrates distributed sensors, image processing, and deep learning models to achieve full automation from data acquisition, status recognition, anomaly detection, and early warning response, reducing manual intervention and making it suitable for intelligent, large-scale microbial culture scenarios. It automatically adjusts the supply of air pressure, steam, circulating water, and usable water to achieve automatic adjustment and optimization of the culture environment, improving the intelligence level and operational convenience of the culture process.

[0130] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program goods. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program goods 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.

[0131] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program goods according to embodiments of this application. 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... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0132] 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.

[0133] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0134] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for honeycomb-type microbial culture and mixing control, characterized in that, include: Install distributed temperature and pressure sensors; Microbial community image data is collected, and Canny edge detection is used to extract the contour information of multiple microorganisms in the image to determine the growth size and maturity of the microorganisms. Based on the convolutional neural network model, the image features are analyzed to determine the current growth size and maturity of the microorganisms, as well as the suitable temperature for microbial culture. Based on the real-time temperature collected at various locations of the microbial community by the distributed temperature sensor, it is determined whether the temperature of the microbial culture is abnormal. When it is determined that the temperature of the microbial culture is abnormal, the circulating water temperature is collected within a preset time period and preprocessed. The preprocessed circulating water temperature is divided into several sliding time windows. Statistical features are extracted for each time window to form a temperature feature vector. Based on the temperature feature vector, it is determined whether the circulating water temperature is abnormal. The circulating water temperature includes the circulating inlet water temperature and the circulating outlet water temperature. When the circulating inlet water temperature is determined to be abnormal, the circulating inlet water temperature is adjusted. Image feature analysis based on convolutional neural network models includes: A historical information dataset was created by combining historical microbial profile information with their corresponding culture temperature thresholds. The historical information dataset is sampled at a ratio of 4:1 to obtain a training subset and a test subset; The neural network model is iteratively trained based on the training subset, the iteratively trained neural network model is evaluated based on the test subset, and it is determined whether to stop iterative training based on the evaluation value. If the evaluation value of the neural network model after the current iteration is less than the evaluation value of the neural network model after the previous iteration, the magnitude of the change in the gradient direction of the neural network model is reduced, and iterative training continues until the preset number of iterations is reached; if the evaluation value of the neural network model after the current iteration is greater than or equal to the evaluation value of the neural network model after the previous iteration, iterative training is stopped, and the convolutional neural network model is obtained.

2. The method for honeycomb microbial culture and mixing control according to claim 1, characterized in that, When using distributed temperature sensors to collect real-time temperatures at various locations within the microbial community to determine if the temperature during microbial culture is abnormal, this includes: A two-dimensional temperature map is established by collecting real-time temperature data from various locations of the microbial community using distributed temperature sensors. An arbitrary temperature data point is selected as the origin in the two-dimensional temperature map, and the correction range is determined by using the origin as the center and the correction radius. The correction temperature is calculated based on the temperature data of each coordinate point within the correction range. When the difference between the temperature data of the origin and the correction temperature is greater than the temperature difference threshold, the origin is determined to be an abnormal data point, and the coordinates of the abnormal data point are recorded. When the abnormal data point exists in the two-dimensional temperature graph, it is determined that the temperature data is abnormal; When there are no abnormal data points in the two-dimensional temperature graph, it is determined that the temperature data is not abnormal.

3. The method for honeycomb microbial culture and mixing control according to claim 2, characterized in that, The collected temperature data is preprocessed, and the preprocessed temperature data is divided into several sliding time windows; Statistical features are extracted for each time window to construct a temperature feature vector, including: The preprocessing includes missing value imputation, normalization, and error value removal; the statistical features include maximum value, minimum value, mean, and standard deviation.

4. The method for honeycomb microbial culture and mixing control according to claim 3, characterized in that, When determining whether the temperature of the circulating water used for microbial culture insulation is abnormal based on temperature feature vectors, the following are also included: Preset circulating water inlet temperature threshold and circulating water outlet temperature threshold; based on the preset circulating water inlet temperature threshold and circulating water outlet temperature threshold; determine whether the circulating water inlet temperature and circulating water outlet temperature are abnormal respectively; When the circulating inlet water temperature is determined to be normal, but the circulating outlet water temperature is abnormal, an early warning is issued, and the warning type and warning level are determined. When it is determined that there are no abnormalities in the circulating water inlet temperature and circulating water outlet temperature at the circulating water interface, an early warning is issued, and the warning type and warning level are determined.

5. The method for honeycomb microbial culture and mixing control according to claim 4, characterized in that, When adjusting the circulating water temperature, the following should be included: Temperature feature vectors are extracted to establish a feature set. The feature set is then integrated with the historical feature set in the historical adjustment set to form a clustering set. The historical adjustment set includes several historical feature sets and several historical adjustment coefficients, and each historical feature set corresponds to a historical adjustment coefficient. All sets to be clustered are processed using K-Means clustering, and the circulating water temperature is adjusted based on the clustering results.

6. The method for honeycomb microbial culture and mixing control according to claim 5, characterized in that, When processing all sets to be clustered using K-Means clustering, and adjusting the circulating water temperature based on the clustering results, the following steps are included: Normalize each data point in the set to be clustered; S1: Initialize K centroids in the set to be clustered, and assign the remaining feature sets to the nearest centroids to form K clusters; S2: Recalculate the centroid of each cluster; S3: Repeat S1 and S2 until the centroid no longer changes; S4: When the cluster containing the feature set does not contain a historical feature set, the historical adjustment coefficient corresponding to the maximum similarity between the feature set and the historical feature set is selected as the initial adjustment coefficient; When the cluster containing the feature set includes a historical feature set, the average value of the historical adjustment coefficients corresponding to all the included historical feature sets is selected as the adjustment coefficient, and the circulating water temperature is adjusted according to the adjustment coefficient.

7. The method for honeycomb microbial culture and mixing control according to claim 6, characterized in that, When selecting the historical adjustment coefficient corresponding to the maximum similarity between the feature set and the historical feature set as the initial adjustment coefficient, it also includes: The average circulating water temperature, the maximum circulating water temperature, and the minimum circulating water temperature are obtained based on the feature set, and the circulating water temperature difference is determined. The circulating water temperature difference is the difference between the maximum circulating water temperature and the average circulating water temperature, or the difference between the minimum circulating water temperature and the average circulating water temperature. When the temperature difference of the circulating water corresponding to the largest absolute value is positive, a first correction coefficient is determined to correct the initial adjustment coefficient, and the circulating water temperature is adjusted with the corrected adjustment coefficient and cultured with the adjusted circulating water temperature. The first correction coefficient is inversely proportional to the temperature difference of the circulating water with the largest absolute value, and the value range of the first correction coefficient is (0.8, 1). When the absolute value of the circulating water temperature difference corresponding to the largest value is negative, a second correction coefficient is determined to correct the initial adjustment coefficient, and the circulating water temperature is adjusted with the corrected adjustment coefficient and cultured with the adjusted circulating water temperature. The second correction coefficient is inversely proportional to the circulating water temperature difference with the largest absolute value, and the value range of the second correction coefficient is (1, 1.2).

8. The method for honeycomb microbial culture and mixing control according to claim 1, characterized in that, Also includes: The pressure value of the gas produced during the growth process of the microorganism is collected, and it is determined whether the pressure value exceeds the pressure threshold. When the pressure value exceeds the pressure threshold, the gas is released to the outside of the microbial growth environment to relieve pressure. Once all the microorganisms have been cultured, they are mixed together. After all the microorganisms have been mixed by rotating and stirring, the microbial mixing is complete.

9. A honeycomb microbial culture and mixing device, used for applying the honeycomb microbial culture and mixing control method as described in any one of claims 1-8, characterized in that, include: A culture tube assembly includes a culture tube body for microbial culture. The culture tube body has a pH sensor interface, a dissolved oxygen sensor interface, a temperature sensor interface, and a pressure sensor interface sequentially arranged on its side. The culture tube body has a water interface, a pressure control interface, a steam interface, and a circulating water interface arranged on its top. A quick-opening pressure cap is also provided on the top of the culture tube body. A one-way pressure valve is provided at the bottom of the culture tube body. A mixing device includes a mixing tank, a stirring device, a motor, and a discharge port; the motor drives the stirring device to rotate and stir; the stirring device is used to stir and mix all microorganisms entering the mixing tank; the mixing tank is provided with at least one culture tube assembly; A water supply device includes a water pump and a water tank; the water tank is provided with an inlet and an outlet at its upper and lower ends, respectively, and the water tank is used to store water for use; the outlet is connected to the water usage interface. The monitoring device includes a pH sensor, a dissolved oxygen sensor, a temperature sensor, and a pressure sensor. Each sensor is connected to the corresponding interface of the pH sensor, dissolved oxygen sensor, temperature sensor, and pressure sensor, respectively, for monitoring and collecting various data within the culture tube body. The control device is electrically connected to the water interface, air pressure control interface, steam interface and circulating water interface, and is used to control the opening degree of each interface; it is also electrically connected to the monitoring device and is used to receive the electrical signals from the monitoring device.

Images