Beer fermentation optimization control method and system based on digital twinning

Abstract

The invention relates to a beer fermentation optimization control method, system and equipment based on digital twinning and a medium. The method comprises the following steps: acquiring beer fermentation detection information; inputting the beer fermentation detection image into the classification neural network model to generate pre-fermentation stage process classification information; according to the process classification information of the pre-fermentation stage, determining a corresponding beer fermentation digital twin sub-model; and adjusting beer fermentation temperature detection information and beer fermentation component detection information based on the beer fermentation digital twin sub-model. By adopting the method, different primary fermentation stages such as a foaming stage, a high-foaming stage, a foaming falling stage and a bubble cap stage can be identified, a digital twinborn model capable of more accurately reflecting and controlling biochemical reaction in the fermentation process can be constructed, the beer fermentation process can be comprehensively monitored, key parameters in the fermentation process can be monitored and analyzed in real time, and the beer fermentation efficiency can be improved. The controllability and the stability of the fermentation process are improved.

Description

Technical Field

[0001] This application belongs to the technical field of beer production, and particularly relates to an optimization control method and system for beer fermentation based on digital twin. Background Art

[0002] With the development of electronic information technology, digital twin technology digitizes complex systems and processes in the physical world by constructing highly accurate simulation models, enabling precise simulation and prediction of their behaviors and performances. It can detect potential problems in advance, optimize design schemes, and monitor equipment status and production parameters in real time, predicting equipment failures, thereby greatly improving manufacturing efficiency and ensuring the stability and reliability of product quality.

[0003] In traditional beer brewing technologies, most rely on the long-term accumulated experience and manual operations of brewing workers. Not only is the degree of manual intervention high, and uncertainties are easily introduced due to fluctuations in the workers' states, but also the risk of misoperations remains high. It is difficult to accurately conduct refined management of the beer fermentation process according to each stage of beer fermentation, and it is even more difficult to meet the growing market demands and quality standards. Summary of the Invention

[0004] Based on this, in view of the above technical problems, it is necessary to provide an optimization control method, system, computer device, and storage medium for beer fermentation based on digital twin that can achieve refined management of the beer fermentation process based on digital twin technology.

[0005] In the first aspect, this application provides an optimization control method for beer fermentation based on digital twin, including:

[0006] Obtain beer fermentation detection information, where the beer fermentation detection information includes beer fermentation detection images, beer fermentation temperature detection information, and beer fermentation component detection information;

[0007] Input the beer fermentation detection image into a classification neural network model to generate pre-fermentation stage process classification information. The pre-fermentation stage process classification information includes foaming period classification information, high-foaming period classification information, falling-foaming period classification information, and foam-cover period classification information. The pre-fermentation stage process classification information is used to distinguish sub-processes of different beer pre-fermentation stages with different biochemical reaction mechanisms;

[0008] According to the pre-fermentation stage process classification information, determine the corresponding beer fermentation digital twin sub-model. Each beer fermentation digital twin sub-model is constructed based on the fermentation equipment used in different sub-processes of the pre-fermentation stage and the temperature and component attributes of the beer fermentation liquid in the fermentation equipment;

[0009] Adjust the beer fermentation temperature detection information and beer fermentation component detection information based on the beer fermentation digital twin sub-model.

[0010] In one embodiment, the classification neural network model is an improved Transformer model, which is obtained by replacing the spatial attention module in the traditional Transformer model with a spatial reduction multi-head attention module;

[0011] The expression of the classification loss function adopted by the improved Transformer model is as follows:

[0012]

[0013] In the formula, L N is the classification loss function, N is the number of categories, α is the discrimination parameter, is the predicted probability of the i-th category, is the true label of the i-th category.

[0014] In one embodiment, the process classification information in the pre-fermentation stage includes the classification information at the end of the beer pre-fermentation stage, and the method further includes:

[0015] Obtain the classification information of beer post-fermentation, which is used to characterize the process method adopted in the beer post-fermentation stage;

[0016] In response to the generation of the classification information at the end of the beer pre-fermentation stage, update the beer fermentation digital twin model to the digital twin model of the operation at the end of the pre-fermentation;

[0017] Based on the digital twin model of the operation at the end of the pre-fermentation, transfer the beer fermentation liquid in the beer pre-fermentation tank to the beer post-fermentation tank for beer post-fermentation;

[0018] According to the classification information of beer post-fermentation, determine the corresponding digital twin model of the post-fermentation stage, and control the beer fermentation temperature detection information and beer fermentation component detection information corresponding to the beer post-fermentation stage based on the digital twin model of the post-fermentation stage.

[0019] In one embodiment, adjusting the beer fermentation temperature detection information and beer fermentation component detection information based on the beer fermentation digital twin model includes:

[0020] Obtain the beer fermentation temperature control curve and beer fermentation component control curve corresponding to the beer fermentation digital twin model. The beer fermentation temperature control curve and beer fermentation component control curve are respectively used to differentially control the temperature and biochemical components of the sub-processes in the beer pre-fermentation stage with different biochemical reaction mechanisms;

[0021] Control the temperature of the beer fermentation liquid to conform to the beer fermentation temperature control curve;

[0022] Input the beer fermentation component detection information and the beer fermentation component control curve into a support vector machine for anomaly detection to generate detection result information, which is used to characterize the process state and flavor state of beer fermentation;

[0023] If the detection result information indicates an anomaly, generate a beer fermentation anomaly prompt message.

[0024] In one embodiment, the beer fermentation optimization control method based on digital twin further includes:

[0025] In response to the generation of the beer fermentation anomaly prompt message, obtain a beer fermentation anomaly handling control instruction;

[0026] Execute the beer fermentation anomaly handling control instruction on the beer fermentation digital twin model to obtain a beer fermentation anomaly handling result;

[0027] If the beer fermentation anomaly handling result is normal, control the beer fermentation physical system corresponding to the beer fermentation digital twin model to execute the beer fermentation anomaly handling control instruction;

[0028] If the beer fermentation anomaly handling result is abnormal, generate a beer fermentation anomaly handling alarm message.

[0029] In one embodiment, the beer fermentation digital twin sub-model includes a beer fermentation substance component sub-model, and the expressions of each beer fermentation substance component sub-model are as follows:

[0030]

[0031] In the formula, is the beer fermentation substance component sub-model corresponding to the j-th beer pre-fermentation stage process classification, X k is the yeast cell concentration at time step k, S k is the fermentable sugar concentration at time step k, E k is the ethanol concentration at time step k, T is the discrete time step, α j is the yeast reproduction coefficient corresponding to the j-th beer pre-fermentation stage process classification, μ m is the specific growth rate of yeast, K s is the saturation constant, β j is the fermentable sugar consumption coefficient corresponding to the j-th beer pre-fermentation stage process classification, is the ethanol specific production rate corresponding to the j-th beer pre-fermentation stage process classification, is the yield coefficient of yeast cells on fermentable sugar corresponding to the j-th beer pre-fermentation stage process classification, ω j is the yeast cell maintenance coefficient corresponding to the j-th beer pre-fermentation stage process classification.

[0032] In one embodiment, the digital twin model of beer fermentation includes a sub-model for analyzing the quality of beer fermentation liquid, and the expressions of each sub-model for analyzing the quality of beer fermentation liquid are as follows:

[0033]

[0034] In the formula, is the sub-model for analyzing the quality of beer fermentation liquid corresponding to the j-th process classification in the pre-fermentation stage of beer, t is the fermentation time, M is the number of types of key substances for analyzing the quality of beer fermentation liquid, and ψ m,j is the weight coefficient of the m-th key substance for analyzing the quality of beer fermentation liquid corresponding to the j-th process classification in the pre-fermentation stage of beer, and ζ m,t is the content of the m-th key substance for analyzing the quality of beer fermentation liquid at time t, and Λ m,j,t is the value at time t of the beer fermentation component control curve of the m-th key substance for analyzing the quality of beer fermentation liquid corresponding to the j-th process classification in the pre-fermentation stage of beer.

[0035] In a second aspect, the present application also provides an optimization control system for beer fermentation based on digital twin, including:

[0036] A beer fermentation detection module, configured to obtain beer fermentation detection information, where the beer fermentation detection information includes beer fermentation detection images, beer fermentation temperature detection information, and beer fermentation component detection information;

[0037] A fermentation process classification module, configured to input the beer fermentation detection images into a classification neural network model to generate pre-fermentation stage process classification information, where the pre-fermentation stage process classification information includes foaming period classification information, high-foaming period classification information, falling-foaming period classification information, and foam-cap period classification information, and the pre-fermentation stage process classification information is used to distinguish sub-processes of different beer pre-fermentation stages with different biochemical reaction mechanisms;

[0038] A digital twin update module, configured to determine the corresponding digital twin model of beer fermentation according to the pre-fermentation stage process classification information, and each digital twin model of beer fermentation is constructed based on the fermentation equipment used in the sub-processes of different pre-fermentation stages and the temperature and component attributes of the beer fermentation liquid in the fermentation equipment;

[0039] A beer fermentation control module, configured to adjust the beer fermentation temperature detection information and the beer fermentation component detection information based on the digital twin model of beer fermentation.

[0040] In a third aspect, the present application also provides a computer device, including a memory and a processor, where the memory stores a computer program, and when the processor executes the computer program, the steps of the method described in any item of the first aspect of the present application are implemented.

[0041] Fourthly, the present application also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the method described in any item of the first aspect of the present application are implemented.

[0042] The above-mentioned digital twin-based beer fermentation optimization control method, system, computer device and storage medium can accurately distinguish different sub-processes (foaming period, high-foaming period, falling-foaming period and foam-cap period) in the pre-fermentation stage by obtaining beer fermentation detection information and using a classification neural network model, and can perform precise control according to the characteristics of each sub-process, which helps to reasonably arrange production plans and resource allocation.

[0043] Furthermore, by determining the digital twin sub-model corresponding to the beer fermentation process classification and adjusting the temperature and composition of the beer fermentation liquid according to the digital twin sub-model, the fermentation process can always be in the best state, improving the quality and taste consistency of the beer, enhancing the overall production efficiency and reducing the production cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following will briefly introduce the drawings required for use in the description of the embodiments or related technologies. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0045] Figure 1 Schematic diagram of the application environment of a digital twin-based beer fermentation optimization control method provided by an embodiment of the present application;

[0046] Figure 2 Schematic diagram of the flow of a digital twin-based beer fermentation optimization control method provided by an embodiment of the present application;

[0047] Figure 3 Schematic diagram of the structure of a classification neural network model provided by an embodiment of the present application;

[0048] Figure 4 Schematic diagram of the flow of a digital twin-based beer post-fermentation stage control method provided by an embodiment of the present application;

[0049] Figure 5 Schematic diagram of the flow of a method for controlling the temperature and composition of beer fermentation provided by an embodiment of the present application;

[0050] Figure 6 Schematic diagram of the flow of an exception handling method provided by an embodiment of the present application;

[0051] Figure 7Schematic flowchart of another digital-twin-based optimized control method for beer fermentation provided by an embodiment of the present application;

[0052] Figure 8 Schematic structural diagram of a digital-twin-based optimized control system for beer fermentation provided by an embodiment of the present application. Detailed implementation manners

[0053] In order to make the objectives, technical solutions and advantages of the present application more clear and understandable, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application, and are not used to limit the present application.

[0054] The digital-twin-based optimized control method for beer fermentation provided by the embodiments of the present application can be applied to an application environment as shown in Figure 1 . Among them, the server 101 can communicate with the sensing device 102 and the working device 103 through a communication channel. The sensing device 102 can send the collected sensing information data to the server 101 through the communication channel. After receiving the sensing information data collected by the sensing device 102, the server 101 can generate a control instruction based on the sensing information data, and can send the control instruction to the working device 103 through the communication channel. After receiving the control instruction, the working device 103 can execute the corresponding work operation instruction.

[0055] Among them, the server 101 can be implemented by an independent server or a server cluster composed of multiple servers. The communication protocol adopted by the communication channel can be but is not limited to RS485, TCP / IP, BACnet, serial communication interface standard, OPC, Web Service, SNMP, HTTP, Wi-Fi, Zigbee protocol (Zigbee), Bluetooth (Bluetooth), Z-Wave, cellular data and near field communication (NFC). The sensing device 102 can be but is not limited to a camera device, a temperature sensing device, a pressure sensing device, an electric flavor sensing device, and a spectral sensing and analysis device. The working device 103 can be but is not limited to various pumps, valves, containers, heating devices, machining devices, and gas supply devices.

[0056] In an exemplary embodiment, as shown in Figure 2 , a digital-twin-based optimized control method for beer fermentation is provided. Taking the method applied to the server 101 in Figure 1 as an example, the method includes the following steps S201 to S204. Among them:

[0057] Step S201, obtain the beer fermentation detection information.

[0058] Specifically, the server 101 may obtain beer fermentation detection information based on the sensing device. The beer fermentation detection information may include beer fermentation detection images, beer fermentation temperature detection information, and beer fermentation component detection information.

[0059] Optionally, the server 101 may obtain beer fermentation detection images of the beer fermentation liquid in the beer fermentation device via a camera built into the beer fermentation device. The beer fermentation detection images may be used to characterize the color, texture, and fermentation degree of the beer fermentation liquid, and thus may reflect the subprocesses of the beer pre-fermentation stage where the beer fermentation liquid is located. The server 101 may obtain beer fermentation temperature detection information via a temperature sensing device. The beer fermentation temperature detection information may include, but is not limited to, the temperature of the beer fermentation liquid and the temperature change rate. The server 101 may obtain beer fermentation component detection information via a component analysis sensing device. The beer fermentation component detection information may include, but is not limited to, the yeast count, fermentable sugar concentration, ethanol concentration, pH value, sediment content, and flavor substance content.

[0060] Step S202: Input the beer fermentation detection images into a classification neural network model to generate pre-fermentation stage process classification information.

[0061] Specifically, the server 101 may input the beer fermentation detection images into a classification neural network model to generate pre-fermentation stage process classification information. The pre-fermentation stage process classification information includes foaming period classification information, high-foaming period classification information, falling-foaming period classification information, and foam-cap period classification information. The pre-fermentation stage process classification information is used to distinguish the subprocesses of the beer pre-fermentation stage with different biochemical reaction mechanisms.

[0062] Optionally, the classification neural network model may be, but is not limited to, constructed based on a multi-layer perceptron, convolutional neural network, recurrent neural network, long short-term memory network, deep residual network, Transformer, YOLO, VGG, or InceptionV4.

[0063] Step S203: Determine the corresponding beer fermentation digital twin model according to the pre-fermentation stage process classification information.

[0064] Specifically, the server 101 may determine the beer fermentation digital twin model corresponding to the pre-fermentation stage process classification information according to the pre-fermentation stage process classification information obtained based on the classification neural network model. Each beer fermentation digital twin model is constructed based on the fermentation equipment used in different subprocesses of the pre-fermentation stage and the temperature and component attributes of the beer fermentation liquid in the fermentation equipment.

[0065] Optionally, the feasible expression of each beer fermentation digital twin model is:

[0066]

[0067] In the formula, Φ j is the j-th beer fermentation digital twin model, is the physical space sub-model of the fermentation tank in the pre-fermentation stage of beer, is the molecular model of beer fermentation substances corresponding to the process classification in the j-th pre-fermentation stage of beer, is the quality analysis sub-model of beer fermentation liquid corresponding to the process classification in the j-th pre-fermentation stage of beer, and t is the moment of beer fermentation, is the temperature sub-model of beer fermentation liquid corresponding to the process classification in the j-th pre-fermentation stage of beer.

[0068] Step S204, adjust the beer fermentation temperature detection information and beer fermentation component detection information based on the beer fermentation digital twin model.

[0069] Specifically, the server 101 can control the working equipment to adjust the beer fermentation temperature detection information and beer fermentation component detection information based on the beer fermentation digital twin model corresponding to the pre-fermentation stage process classification information.

[0070] Optionally, the working equipment may include but is not limited to material feeding equipment, stirring equipment, heating equipment, cooling equipment, and ventilation equipment.

[0071] Illustratively, the server 101 can accurately control conditions such as the initial temperature of beer fermentation, yeast type, yeast inoculation amount, and fermentable sugar concentration according to the beer fermentation digital twin model during the foaming period, ensuring the growth and reproduction of yeast and ensuring the smooth start of beer fermentation; the server 101 can control the temperature change rate of beer fermentation liquid, fermentation temperature range, and nutrient component supply during the high-foaming period according to the beer fermentation digital twin model during the high-foaming period, promoting efficient yeast metabolism and the synthesis of alcohol and flavor substances; the server 101 can reasonably adjust the temperature, acidity, and concentration of beer fermentation liquid according to the beer fermentation digital twin model during the falling-foam period to promote yeast precipitation; the server 101 can strictly monitor the temperature and flavor substance components of beer fermentation liquid according to the beer fermentation digital twin model during the foam-cap period to ensure the quality in the later stage of fermentation.

[0072] In the above beer fermentation optimization control method based on digital twin, by obtaining beer fermentation detection information including image, temperature, and component information, accurately dividing the sub-processes in the pre-fermentation stage with the help of a classification neural network model, determining the digital twin model corresponding to the beer fermentation process classification, and adjusting the temperature and components of beer fermentation liquid according to the digital twin model, refined management of each sub-process in the beer pre-fermentation stage can be realized. Furthermore, the beer fermentation process can be comprehensively monitored to ensure that key parameters in the fermentation process are monitored and analyzed in real time, improving the controllability and stability of the fermentation process.

[0073] In an alternative embodiment, as Figure 3 shown, the classification neural network model can be an improved Transformer model, and the improved Transformer model can be obtained by replacing the spatial attention module in the traditional Transformer model with a spatial reduction multi-head attention module.

[0074] Optionally, the improved Transformer model can be sequentially connected with a cross-scale attention module, a residual convolution module, and a global attention module after the Transformer, which can adaptively weight and fuse features of different scales, highlight the features more valuable for the classification task, suppress noise and irrelevant information, thereby improving the accuracy and reliability of classification.

[0075] A feasible expression of the classification loss function adopted by the improved Transformer model is as follows:

[0076]

[0077] In the formula, L N is the classification loss function, N is the number of categories, α is the discriminant parameter, is the predicted probability of the i-th category, is the true label of the i-th category.

[0078] In the above digital twin-based beer fermentation optimization control method, by replacing the spatial attention module in the traditional Transformer model with a spatial reduction multi-head attention module, the amount of calculation and the number of model parameters can be effectively reduced. When processing data such as beer fermentation detection images, due to the large amount of image data, this improvement can accelerate the training and inference speed of the model. The design of this classification loss function helps the model converge faster during training, can find a better classification decision boundary, and further enables the model to better adapt to the characteristics of different beer fermentation data sets, and can improve the generalization ability of the model.

[0079] In an alternative embodiment, the process classification information in the pre-fermentation stage may include the classification information at the end of the beer pre-fermentation stage. Please refer to Figure 4 , and the digital twin-based beer fermentation optimization control method further includes:

[0080] Step S401, obtaining the classification information of the beer post-fermentation.

[0081] Specifically, to obtain the classification information of the beer post-fermentation, the classification information of the beer post-fermentation can be used to characterize the process method adopted in the beer post-fermentation stage.

[0082] Step S402: In response to the generation of the classification information at the end of the beer pre-fermentation stage, update the beer fermentation digital twin model to the operation digital twin model at the end of pre-fermentation.

[0083] Schematically, the operation digital twin model at the end of pre-fermentation may include, but is not limited to, the digital twin component of the fermentation tank, the digital twin component of the fermentation material transportation pipeline network, and the digital twin component of the dynamic transportation of the fermented material.

[0084] Step S403: Transfer the beer fermentation liquid in the beer pre-fermentation tank to the beer post-fermentation tank for beer post-fermentation based on the operation digital twin model at the end of pre-fermentation.

[0085] Schematically, the transfer of the beer fermentation liquid in the beer pre-fermentation tank to the beer post-fermentation tank can be controlled based on the valve digital twin model in the digital twin component of the fermentation material transportation pipeline network, and the transportation process of the beer fermentation material can be simulated and monitored based on the digital twin component of the dynamic transportation of the fermented material and the digital twin component of the fermentation tank.

[0086] Step S404: Determine the corresponding digital twin model of the post-fermentation stage according to the beer post-fermentation classification information, and control the beer fermentation temperature detection information and the beer fermentation component detection information corresponding to the beer post-fermentation stage based on the digital twin model of the post-fermentation stage.

[0087] In the above digital twin-based beer fermentation optimization control method, through the operation digital twin model at the end of pre-fermentation, the fermentation liquid can be accurately transferred to the post-fermentation tank according to the actual state at the end of pre-fermentation, ensuring the seamless connection between the pre-fermentation and post-fermentation processes, thereby avoiding the quality fluctuation of the fermentation liquid caused by improper transfer process, providing good starting conditions for the post-fermentation stage, and helping to maintain the stability and continuity of the whole fermentation process.

[0088] Furthermore, the corresponding digital twin model of the post-fermentation stage can be determined according to different beer post-fermentation classification information, and then the beer fermentation temperature detection information and the beer fermentation component detection information in the post-fermentation stage can be accurately controlled based on this model. It can ensure the slow fermentation of the residual sugar, the delicate conversion of flavor substances, and the clarification and maturation of the beer during the post-fermentation process, thereby improving the quality and taste complexity of the beer and ensuring the consistency of the beer quality.

[0089] In an optional embodiment, please refer to Figure 5 , adjusting the beer fermentation temperature detection information and the beer fermentation component detection information based on the beer fermentation digital twin model includes:

[0090] Step S501: Obtain the beer fermentation temperature control curve and the beer fermentation component control curve corresponding to the beer fermentation digital twin model.

[0091] Specifically, the beer fermentation temperature control curve and the beer fermentation component control curve corresponding to each beer fermentation digital twin model can be obtained. The beer fermentation temperature control curve and the beer fermentation component control curve can be used to differentially control the temperature and biochemical components of the sub-processes in the pre-fermentation stage of beer with different biochemical reaction mechanisms respectively.

[0092] Step S502: Control the temperature of the beer fermentation broth to conform to the beer fermentation temperature control curve.

[0093] Specifically, the server can calculate the error between the real-time temperature of the beer fermentation broth and the beer fermentation temperature control curve corresponding to the beer fermentation digital twin model. When the error exceeds the preset temperature threshold, the error is controlled to be less than the preset temperature control threshold through heating equipment or cooling equipment.

[0094] Step S503: Input the beer fermentation component detection information and the beer fermentation component control curve into a support vector machine for anomaly detection to generate detection result information.

[0095] Specifically, the beer fermentation component detection information and the beer fermentation component control curve can be input into a support vector machine for anomaly detection to generate detection result information, and the detection result information can be used to characterize the process state and flavor state of beer fermentation.

[0096] Exemplarily, a feasible expression of the support vector machine is as follows:

[0097]

[0098] In the formula, L max (α) is the objective optimization function term of the support vector machine, α is the Lagrange multiplier and α i ∈[0,C] and where C is the regularization parameter, is the feature vector, y i ,y j is the class label corresponding to the feature vector, is the kernel function, is the normal vector of the hyperplane, b is the bias term of the hyperplane, is the feature vector to the distance of the hyperplane.

[0099] Step S504: If the detection result information indicates an anomaly, generate a beer fermentation anomaly prompt message.

[0100] In the above-mentioned beer fermentation optimization control method based on digital twin, by obtaining the temperature and composition control curves corresponding to the digital twin sub-models of beer fermentation, precise temperature and composition regulation can be implemented for the unique biochemical reaction mechanisms of each sub-process in the pre-fermentation stage. At the same time, the beer fermentation component detection information and the composition control curve are input into the support vector machine for anomaly detection, and the generated detection result information covers the process status and flavor status of beer fermentation. It can not only understand whether the beer fermentation meets the expected progress, but also predict in advance the development trend of beer flavor, realizing comprehensive and real-time monitoring of the beer fermentation process, and further ensuring that the fermentation process is always in the best state, reducing delays and waste caused by abnormal fermentation.

[0101] In an alternative embodiment, please refer to Figure 6 , the beer fermentation optimization control method based on digital twin further includes:

[0102] Step S601, in response to the generation of beer fermentation anomaly prompt information, obtain the beer fermentation anomaly handling control instruction.

[0103] Illustratively, the acquisition channels of the beer fermentation anomaly handling control instruction may include but are not limited to obtaining the beer fermentation anomaly handling control instruction input by the staff through the human-computer interaction device, analyzing and obtaining based on big data and artificial intelligence algorithms, and analyzing and obtaining based on the digital twin model and preset processing rules.

[0104] Step S602, execute the beer fermentation anomaly handling control instruction on the beer fermentation digital twin model to obtain the beer fermentation anomaly handling result.

[0105] Step S603, if the beer fermentation anomaly handling result is normal, control the beer fermentation physical system corresponding to the beer fermentation digital twin model to execute the beer fermentation anomaly handling control instruction.

[0106] Step S604, if the beer fermentation anomaly handling result is abnormal, generate the beer fermentation anomaly handling alarm information.

[0107] In the above-mentioned beer fermentation optimization control method based on digital twin, based on the anomaly handling control instruction, the anomaly response measures can be quickly started, avoiding the deterioration of abnormal fermentation caused by untimely problem discovery, and thus reducing the possible product quality loss and production delay. By first executing the anomaly handling control instruction on the beer fermentation digital twin model, using the model's simulation and prediction capabilities of the fermentation process, the effects of the treatment plan can be evaluated in advance. It helps to select the optimal treatment strategy to ensure that problems can be solved more effectively when implemented in the actual production equipment, and minimizes the possibility of production interruption or further failures caused by incorrect handling.

[0108] In an alternative embodiment, the digital twin model of beer fermentation includes a molecular model of beer fermentation substances, and the feasible expressions of each molecular model of beer fermentation substances are as follows:

[0109]

[0110] Wherein, is the molecular model of beer fermentation substances corresponding to the j-th process classification of the pre-fermentation stage of beer, X k is the yeast cell concentration at time k, S k is the fermentable sugar concentration at time k, E k is the ethanol concentration at time k, T is the discrete time step, α j is the yeast reproduction coefficient corresponding to the j-th process classification of the pre-fermentation stage of beer, μ m is the specific growth rate of yeast, K s is the saturation constant, β j is the fermentable sugar consumption coefficient corresponding to the j-th process classification of the pre-fermentation stage of beer, is the specific ethanol production rate corresponding to the j-th process classification of the pre-fermentation stage of beer, is the yield coefficient of yeast cells on fermentable sugar corresponding to the j-th process classification of the pre-fermentation stage of beer, ω j is the yeast cell maintenance coefficient corresponding to the j-th process classification of the pre-fermentation stage of beer.

[0111] Schematically, in the digital twin model of beer fermentation during the foaming period, the yeast reproduction coefficient can be higher than that of other digital twin models of beer fermentation. In the digital twin model of beer fermentation during the high-foaming period, the fermentable sugar consumption coefficient and the specific ethanol production rate can be higher than those of other digital twin models of beer fermentation.

[0112] In the above method for optimizing the control of beer fermentation based on digital twin, the input and utilization of resources can be optimized based on the accurate prediction of the changes in fermentation substances by the digital twin model.

[0113] In an alternative embodiment, the digital twin model of beer fermentation includes a sub-model for analyzing the quality of beer fermentation broth, and the expressions of each sub-model for analyzing the quality of beer fermentation broth are as follows:

[0114]

[0115] Wherein, is the sub-model for analyzing the quality of beer fermentation broth corresponding to the j-th process classification of the pre-fermentation stage of beer, t is the fermentation time, M is the number of types of key substances for analyzing the quality of beer fermentation broth, ψ m,j is the weight coefficient of the i-th key substance for analyzing the quality of beer fermentation broth corresponding to the j-th process classification of the pre-fermentation stage of beer, ζ m,tis the content of the m-th key substance for the quality analysis of the beer fermentation liquid at time t, Λ m,j,t is the value at time t of the beer fermentation component control curve of the m-th key substance for the quality analysis of the beer fermentation liquid corresponding to the j-th process classification in the pre-fermentation stage of the beer.

[0116] Furthermore, the expression of each sub-model for the quality analysis of the beer fermentation liquid can be as follows:

[0117]

[0118] In the formula, M j is the number of types of key substances for the quality analysis of the beer fermentation liquid corresponding to the j-th process classification in the pre-fermentation stage of the beer.

[0119] In the above beer fermentation optimization control method based on digital twin, it is possible to comprehensively consider the influence of multiple key substances for the quality analysis of the beer fermentation liquid on the beer quality based on the sub-model for the quality analysis of the beer fermentation liquid, and comprehensively evaluate the quality of the fermentation liquid.

[0120] In an optional embodiment, please refer to Figure 7 , the beer fermentation optimization control method based on digital twin may include:

[0121] Step S701, obtain the beer fermentation detection information.

[0122] Step S702, input the beer fermentation detection image into the classification neural network model to generate the process classification information for the pre-fermentation stage.

[0123] Step S703, determine the corresponding digital twin sub-model for the beer fermentation according to the process classification information for the pre-fermentation stage.

[0124] Step S704, obtain the beer fermentation temperature control curve and the beer fermentation component control curve corresponding to the digital twin sub-model for the beer fermentation.

[0125] Step S705, control the temperature of the beer fermentation liquid to conform to the beer fermentation temperature control curve.

[0126] Step S706, input the beer fermentation component detection information and the beer fermentation component control curve into the support vector machine for anomaly detection to generate the detection result information.

[0127] Step S707, if the detection result information indicates an anomaly, generate a beer fermentation anomaly prompt message.

[0128] Step S708, in response to the generation of the beer fermentation anomaly prompt message, obtain the beer fermentation anomaly handling control instruction.

[0129] Step S709: Execute the beer fermentation abnormal handling control instruction on the beer fermentation digital twin model to obtain the beer fermentation abnormal handling result.

[0130] Step S710: If the beer fermentation abnormal handling result is normal, control the beer fermentation physical system corresponding to the beer fermentation digital twin model to execute the beer fermentation abnormal handling control instruction.

[0131] Step S711: If the beer fermentation abnormal handling result is abnormal, generate a beer fermentation abnormal handling alarm message.

[0132] In the above beer fermentation optimization control method based on digital twins, the temperature and composition of the beer fermentation liquid can be precisely controlled according to the temperature control curve and composition control curve obtained from the beer fermentation digital twin sub-model. After obtaining the beer fermentation abnormal handling control instruction, the instruction can be first executed on the beer fermentation digital twin model to obtain the abnormal handling result. The effectiveness and feasibility of the handling scheme can be evaluated without affecting the actual fermentation process.

[0133] It should be understood that although the steps in the flowcharts of the various embodiments described above are shown in sequence according to the arrows, these steps do not necessarily have to be executed in the order indicated by the arrows. Unless there is a clear indication in this article, the execution of these steps is not strictly limited in order, and these steps can be executed in other orders. Moreover, at least some of the steps in the flowcharts of the various embodiments described above may include multiple steps or multiple stages. These steps or stages do not necessarily have to be executed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be executed alternately or in turn with at least a part of other steps or steps in other steps.

[0134] Based on the same inventive concept, an embodiment of the present application also provides a beer fermentation optimization control system based on digital twins for implementing the above-mentioned beer fermentation optimization control method based on digital twins. The solution provided by this system to solve the problem is similar to the solution described in the above method. Therefore, the specific limitations in one or more embodiments of the beer fermentation optimization control system based on digital twins provided below can refer to the limitations on the beer fermentation optimization control method based on digital twins in the above text, and will not be repeated here.

[0135] In an exemplary embodiment, as Figure 8 shown, a beer fermentation optimization control system 800 based on digital twins is provided, including:

[0136] The beer fermentation detection module 801 is used to obtain beer fermentation detection information, where the beer fermentation detection information includes beer fermentation detection images, beer fermentation temperature detection information, and beer fermentation component detection information;

[0137] The fermentation process classification module 802 is used to input the beer fermentation detection images into a classification neural network model to generate pre-fermentation stage process classification information. The pre-fermentation stage process classification information includes foaming period classification information, high-foaming period classification information, foam-collapse period classification information, and foam-cap period classification information. The pre-fermentation stage process classification information is used to distinguish the sub-processes of the beer pre-fermentation stage with different biochemical reaction mechanisms;

[0138] The digital twin update module 803 is used to determine the corresponding beer fermentation digital twin sub-models according to the pre-fermentation stage process classification information. Each beer fermentation digital twin sub-model is constructed based on the fermentation equipment used in the sub-processes of different pre-fermentation stages and the temperature and component attributes of the beer fermentation liquid in the fermentation equipment;

[0139] The beer fermentation control module 804 is used to adjust the beer fermentation temperature detection information and the beer fermentation component detection information based on the beer fermentation digital twin sub-models.

[0140] In an alternative embodiment, please refer to Figure 8 , the beer fermentation optimization control system 800 based on digital twin can also be used for:

[0141] Obtain beer post-fermentation classification information, where the beer post-fermentation classification information is used to characterize the process method adopted in the beer post-fermentation stage;

[0142] In response to the generation of the beer pre-fermentation stage end classification information, update the beer fermentation digital twin sub-model to the pre-fermentation end operation digital twin sub-model, and transfer the beer fermentation liquid in the beer pre-fermentation tank to the beer post-fermentation tank based on the pre-fermentation end operation digital twin sub-model for beer post-fermentation;

[0143] Determine the corresponding post-fermentation stage digital twin sub-model according to the beer post-fermentation classification information;

[0144] Control the beer fermentation temperature detection information and the beer fermentation component detection information corresponding to the beer post-fermentation stage based on the post-fermentation stage digital twin sub-model.

[0145] In an alternative embodiment, please refer to Figure 8 , the beer fermentation control module 804 can also be used for:

[0146] Obtain the beer fermentation temperature control curve and the beer fermentation component control curve corresponding to the beer fermentation digital twin model. The beer fermentation temperature control curve and the beer fermentation component control curve are respectively used to differentially control the temperature and biochemical components of the sub-processes in the pre-fermentation stage of beer with different biochemical reaction mechanisms;

[0147] Control the temperature of the beer fermentation liquid to conform to the beer fermentation temperature control curve;

[0148] Input the beer fermentation component detection information and the beer fermentation component control curve into a support vector machine for anomaly detection to generate detection result information, which is used to characterize the process state and flavor state of beer fermentation;

[0149] If the detection result information indicates an anomaly, generate a beer fermentation anomaly prompt message.

[0150] In an optional embodiment, please refer to Figure 8 , the beer fermentation control module 804 can also be used for:

[0151] In response to the generation of the beer fermentation anomaly prompt message, obtain a beer fermentation anomaly handling control instruction;

[0152] Execute the beer fermentation anomaly handling control instruction on the beer fermentation digital twin model to obtain a beer fermentation anomaly handling result;

[0153] If the beer fermentation anomaly handling result is normal, control the beer fermentation physical system corresponding to the beer fermentation digital twin model to execute the beer fermentation anomaly handling control instruction;

[0154] If the beer fermentation anomaly handling result is abnormal, generate a beer fermentation anomaly handling alarm message.

[0155] In one embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the steps of a power supply safety management method as described above are implemented.

[0156] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0157] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to the partial descriptions of the method embodiments. The device embodiments described above are only illustrative. The components described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the present disclosure solution. Those of ordinary skill in the art can understand and implement it without creative efforts.

[0158] The above embodiments only represent several implementation manners of the embodiments of the present application. The description is relatively specific and detailed, but it should not be construed as a limitation on the scope of the patent for the embodiments of the application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the embodiments of the present application, several modifications and improvements can still be made, and these all belong to the protection scope of the embodiments of the present application.

Claims

1. A method for optimizing the control of beer fermentation based on digital twins, characterized in that, The method includes: Obtaining beer fermentation detection information, where the beer fermentation detection information includes a beer fermentation detection image, beer fermentation temperature detection information, and beer fermentation component detection information; Inputting the beer fermentation detection image into a classification neural network model to generate pre-fermentation stage process classification information, where the pre-fermentation stage process classification information includes foaming period classification information, high-foaming period classification information, foam-collapse period classification information, and foam-cap period classification information, and the pre-fermentation stage process classification information is used to distinguish sub-processes of different beer pre-fermentation stages with different biochemical reaction mechanisms; According to the pre-fermentation stage process classification information, determining the corresponding beer fermentation digital twin models, where each of the beer fermentation digital twin models is constructed based on the fermentation equipment used in different sub-processes of the pre-fermentation stage and the temperature and component attributes of the beer fermentation liquid in the fermentation equipment; Adjusting the beer fermentation temperature detection information and the beer fermentation component detection information based on the beer fermentation digital twin models.

2. The method according to claim 1, characterized in that, The classification neural network model is an improved Transformer model, and the improved Transformer model is obtained by replacing the spatial attention module in the traditional Transformer model with a spatial reduction multi-head attention module; The expression of the classification loss function adopted by the improved Transformer model is as follows: where L N is the classification loss function, N is the number of classes, α is the discriminant parameter, is the predicted probability for the i-th class, is the true label of the i-th class.

3. The method according to claim 1, characterized in that The pre-fermentation stage process classification information includes beer pre-fermentation stage end classification information, and the method further includes: Obtaining beer post-fermentation classification information, where the beer post-fermentation classification information is used to characterize the process method adopted in the beer post-fermentation stage; In response to the generation of the beer pre-fermentation stage end classification information, updating the beer fermentation digital twin model to a pre-fermentation end operation digital twin model; Transferring the beer fermentation liquid in the beer pre-fermentation tank to the beer post-fermentation tank for beer post-fermentation based on the pre-fermentation end operation digital twin model; According to the beer post-fermentation classification information, determining the corresponding post-fermentation stage digital twin model, and controlling the beer fermentation temperature detection information and the beer fermentation component detection information corresponding to the beer post-fermentation stage based on the post-fermentation stage digital twin model.

4. The method according to claim 1, wherein The adjusting the beer fermentation temperature detection information and the beer fermentation component detection information based on the beer fermentation digital twin models includes: Obtaining a beer fermentation temperature control curve and a beer fermentation component control curve corresponding to the beer fermentation digital twin models, where the beer fermentation temperature control curve and the beer fermentation component control curve are respectively used to differentially control the temperature and biochemical components of different sub-processes of the beer pre-fermentation stage with different biochemical reaction mechanisms; Controlling the temperature of the beer fermentation liquid to conform to the beer fermentation temperature control curve; Inputting the beer fermentation component detection information and the beer fermentation component control curve into a support vector machine for anomaly detection to generate detection result information, where the detection result information is used to characterize the process state and flavor state of beer fermentation; If the detected result information is abnormal, generate a beer fermentation abnormality prompt message.

5. The method according to claim 4, characterized in that The method further includes: In response to the generation of the beer fermentation abnormality prompt message, obtain a beer fermentation abnormality handling control instruction; Execute the beer fermentation abnormality handling control instruction on the beer fermentation digital twin model to obtain a beer fermentation abnormality handling result; If the beer fermentation abnormality handling result is normal, control the beer fermentation physical system corresponding to the beer fermentation digital twin model to execute the beer fermentation abnormality handling control instruction; If the beer fermentation abnormality handling result is abnormal, generate a beer fermentation abnormality handling alarm message.

6. The method according to any one of claims 1 to 5, characterized in that, The beer fermentation digital twin sub-model includes a beer fermentation material component sub-model, and the expressions of each beer fermentation material component sub-model are as follows: Wherein, is the molecular model of beer fermentation substances corresponding to the j-th process classification in the pre-fermentation stage of beer, X k is the yeast cell concentration at time k, S k is the concentration of fermentable sugar at time k, E k is the ethanol concentration at time k, T is the discrete time step, α j is the yeast propagation coefficient corresponding to the j-th process classification in the pre-fermentation stage of beer, μ m is the specific growth rate of yeast, K s is the saturation constant, β j is the fermentable sugar consumption coefficient corresponding to the j-th process classification in the pre-fermentation stage of beer, is the specific ethanol production rate corresponding to the j-th process classification in the pre-fermentation stage of beer, is the yield coefficient of yeast cells on fermentable sugar corresponding to the j-th process classification in the pre-fermentation stage of beer, ω j is the maintenance coefficient of yeast cells corresponding to the j-th process classification in the pre-fermentation stage of beer.

7. The method according to claim 6, characterized in that The beer fermentation digital twin sub-model includes a beer fermentation liquid quality analysis sub-model, and the expressions of each beer fermentation liquid quality analysis sub-model are as follows: In the formula, is the sub-model for analyzing the quality of beer fermentation broth corresponding to the j-th process classification in the pre-fermentation stage of beer, t is the fermentation time, M is the number of types of key substances for analyzing the quality of beer fermentation broth, and ψ m,j is the weight coefficient of the m-th key substance for analyzing the quality of beer fermentation broth corresponding to the j-th process classification in the pre-fermentation stage of beer, and ζ m,t is the content of the m-th key substance for analyzing the quality of beer fermentation broth at time t, and Λ m,j,t is the value at time t of the control curve for beer fermentation components of the m-th key substance for analyzing the quality of beer fermentation broth corresponding to the j-th process classification in the pre-fermentation stage of beer.

8. An optimized control system for beer fermentation based on digital twin, characterized in that, The system includes: A beer fermentation detection module, configured to obtain beer fermentation detection information, where the beer fermentation detection information includes a beer fermentation detection image, beer fermentation temperature detection information, and beer fermentation component detection information; A fermentation process classification module, configured to input the beer fermentation detection image into a classification neural network model to generate pre-fermentation stage process classification information, where the pre-fermentation stage process classification information includes foaming period classification information, high-foaming period classification information, foam-collapse period classification information, and foam-cap period classification information, and the pre-fermentation stage process classification information is used to distinguish sub-processes of different beer pre-fermentation stages with different biochemical reaction mechanisms; A digital twin update module, configured to determine a corresponding beer fermentation digital twin sub-model according to the pre-fermentation stage process classification information, and each beer fermentation digital twin sub-model is constructed based on the fermentation equipment used in different sub-processes of the pre-fermentation stage and the temperature and component attributes of the beer fermentation liquid in the fermentation equipment; A beer fermentation control module, configured to adjust the beer fermentation temperature detection information and the beer fermentation component detection information based on the beer fermentation digital twin model.

9. A computer device, comprising a memory and a processor, the memory storing a computer program, characterized in that, When the processor executes the computer program, it implements the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the method according to any one of claims 1 to 7.

Images