Beer fermentation process temperature control method

Abstract

The invention belongs to the technical field of brewing industry, and discloses a beer fermentation process temperature control method, which comprises the following steps: establishing a fermentation knowledge base; obtaining the type of beer to be fermented and corresponding fermentation knowledge; a timer is set, timing is started, and fermentation state parameters and temperature are collected at the same time; determining a current fermentation stage according to the collected fermentation state parameters and stage characteristics; obtaining a corresponding target temperature obtaining model according to the current fermentation stage, and inputting the collected fermentation state parameters into the model to obtain a target temperature; and regulating and controlling according to the fermentation liquid temperature and the target temperature. According to the method, the fermentation stage is accurately judged by calculating the similarity between the current state and the adjacent stage characteristics and introducing the time weight; the non-linear relation between the fermentation state parameters and the optimal control temperature is captured by using a machine learning model, accurate temperature control is realized, the limitation of a traditional method is overcome, and the controllability of the fermentation process and the stability of beer quality are improved.

Description

Technical Field

[0001] The invention relates to the technical field of brewing industry, and more particularly to a temperature control method for beer fermentation process. Background Art

[0002] The fermentation stage in the beer production process plays a decisive role in product quality. With the continuous expansion of the beer market and consumers' increasing attention to quality, how to accurately control the fermentation process to improve beer quality and flavor consistency has become an important issue in the beer production industry.

[0003] Beer fermentation is a complex biochemical process involving multiple stages, each with its own unique characteristics and temperature requirements. Traditional fermentation control methods mainly rely on experience and fixed schedules to determine the fermentation stage and control the temperature. This method has the following problems. First, judging the fermentation stage based solely on time lacks flexibility and accuracy. Due to the influence of factors such as raw materials and yeast activity, the actual fermentation process may deviate from the preset schedule, resulting in inaccurate stage judgment, which in turn affects subsequent temperature control. Secondly, the temperature of each stage is controlled within a fixed range, and it is impossible to make precise adjustments to the dynamic changes of the fermentation process. Fermentation is a dynamic process, and factors such as yeast activity and metabolite concentration will affect the optimal fermentation temperature. A fixed temperature range is difficult to meet this dynamic demand. This in turn affects the metabolic activity of yeast and the production of flavor substances, reducing the quality and flavor of beer. Summary of the invention

[0004] In order to overcome the problem in the prior art that the quality and flavor of beer may be reduced, the present invention proposes a temperature control method for a beer fermentation process to solve the above problem.

[0005] The present invention provides the following technical solutions: A method for controlling temperature in a beer fermentation process, comprising: Establishing a fermentation knowledge base, the fermentation knowledge base includes beer types and corresponding fermentation knowledge; the fermentation knowledge includes fermentation stages, target temperature acquisition models corresponding to the fermentation stages, and stage information; the stage information includes stage duration and stage characteristics; Obtain the type of beer to be fermented, and obtain corresponding fermentation knowledge according to the type of beer to be fermented; Set a timer and start timing when fermentation starts, while collecting fermentation state parameters and fermentation liquid temperature; Determine the current fermentation stage according to the fermentation state parameters and stage characteristics collected recently; According to the determined current fermentation stage, a corresponding target temperature acquisition model is obtained and recorded as the current model; the most recently collected fermentation state parameters are input into the current model to obtain the target temperature; Regulate according to the fermentation liquid temperature and target temperature.

[0006] Preferably, the fermentation stages include, in order: an adaptation stage, a logarithmic growth stage, a stable stage, and a decay stage; the stage characteristics are represented by a characteristic parameter array under each stage, and the characteristic parameter array includes three elements: an average carbon dioxide production rate, an average specific gravity change rate, and an average pH change rate; the stage duration is the maximum duration under each stage; The fermentation state parameters include carbon dioxide production rate, specific gravity change rate, pH change rate, and concentrations of alcohols, esters, and diacetyl.

[0007] Preferably, the step of collecting fermentation state parameters comprises: At preset intervals :Measure the volume of carbon dioxide discharged by flow meter and calculate the carbon dioxide production rate; measure the specific gravity of fermentation liquid by hydrometer and calculate the specific gravity change rate; measure the pH value of fermentation liquid by pH meter and calculate the pH change rate; At preset intervals , obtain fermentation broth samples and determine the concentrations of alcohols, esters and diacetyl using corresponding equipment.

[0008] Preferably, the step of determining the current fermentation stage according to the most recently collected fermentation state parameters and stage characteristics comprises: Check whether there is a determined current fermentation stage, if so, record the determined current fermentation stage as the first fermentation stage, otherwise record the adaptation stage as the first fermentation stage; Check the first fermentation stage for subsequent stages; If not present, the current fermentation stage is determined as the first fermentation stage; If present, the subsequent stage was obtained and recorded as the second fermentation stage; Obtaining stage features corresponding to the first fermentation stage and the second fermentation stage, and respectively calculating a first similarity and a second similarity according to the most recently collected fermentation state parameters; Obtain the duration of the current stage and the stage duration corresponding to the first fermentation stage to obtain a first weight and a second weight; Compare whether the first similarity multiplied by the first weight is greater than or equal to the second similarity multiplied by the second weight; if so, determine the current fermentation stage as the first fermentation stage; if not, determine the current fermentation stage as the second fermentation stage.

[0009] Preferably, the first similarity is calculated by the following formula: , In the formula, represents the first similarity, represents the average carbon dioxide production rate in the stage characteristics corresponding to the first fermentation stage, Indicates the carbon dioxide production rate in the most recently collected fermentation state parameters. It represents the average specific gravity change rate in the stage characteristics corresponding to the first fermentation stage, Indicates the rate of change of specific gravity in the most recently collected fermentation state parameters. represents the average pH change rate in the stage characteristics corresponding to the first fermentation stage, Indicates the pH change rate of the most recently collected fermentation state parameters; The second similarity is calculated by the following formula: , In the formula, represents the second similarity, represents the average carbon dioxide production rate in the stage characteristics corresponding to the second fermentation stage, represents the average specific gravity change rate in the stage characteristics corresponding to the second fermentation stage, represents the average pH change rate in the stage characteristics corresponding to the second fermentation stage; The first weight is expressed by the following formula: , to calculate, where represents the first weight, Indicates the duration of the first fermentation stage, Indicates the duration of the first fermentation stage. represents the preset scaling factor and ; The second weight is calculated by the following formula: , to calculate, where Represents the second weight.

[0010] Preferably, the process of establishing the target temperature acquisition model corresponding to the fermentation stage includes: According to the historical fermentation state parameters and the corresponding fermentation liquid temperature, the learning data set corresponding to each fermentation stage is obtained; Divide the corresponding learning data set into a training set and a validation set according to a preset ratio; Build a machine learning framework and define the input layer and output layer, set the loss function and optimizer, and build a machine learning model; The fermentation state parameters are used as input layer data, and the corresponding fermentation liquid temperature is used as output layer data. The training set is used to train the constructed machine learning model according to the set batch size and number of training rounds. The trained machine learning model is tested using the test set, and a machine learning model that meets the preset accuracy is output as the target temperature acquisition model corresponding to the fermentation stage.

[0011] Preferably, the step of obtaining a learning data set corresponding to each fermentation stage according to the historical fermentation state parameters and the corresponding fermentation liquid temperature comprises: Based on the preset standards, batches whose fermentation results meet the preset standards are selected from the historical fermentation records; For each selected batch, for each fermentation stage in the fermentation process: obtain the fermentation state parameters in the fermentation process at that stage, and obtain the fermentation liquid temperature at the time point corresponding to the fermentation state parameters; and combine the obtained fermentation state parameters and fermentation liquid temperature into a piece of learning data for that stage; Multiple learning data of each fermentation stage are aggregated to form a learning data set corresponding to each fermentation stage.

[0012] Preferably, the regulation according to the fermentation liquid temperature and the target temperature includes: Real-time monitoring of fermentation liquid temperature and target temperature; When the deviation between the fermentation liquid temperature and the target temperature exceeds a preset threshold, the power of the temperature control device is adjusted accordingly until the fermentation liquid temperature reaches the target temperature; and when the fermentation liquid temperature reaches the target temperature value, the temperature control device is set to the insulation mode.

[0013] The present invention provides a method for controlling the temperature of a beer fermentation process, which has the following beneficial effects: By collecting the carbon dioxide production rate, specific gravity change rate, and pH change rate to represent the current state characteristics, and then calculating the similarity between the current state and the characteristics of the two adjacent stages, the time weight is introduced to weight the similarity, and the current stage is judged based on the weighted similarity. This overcomes the limitations of traditional judgment that relies solely on time and significantly improves the accuracy of fermentation stage judgment.

[0014] By establishing a machine learning model as the target temperature acquisition model, taking the fermentation state parameters as input and the target temperature as output, the complex nonlinear relationship between the fermentation state parameters and the optimal control temperature is captured. According to the current fermentation stage, the corresponding target temperature acquisition model is called, and the recently collected fermentation state parameters are input to obtain the precise target temperature. Then, by real-time monitoring of the deviation between the fermentation liquid temperature and the target temperature, the PID control algorithm is used to dynamically adjust the power of the temperature control equipment to achieve fast response and stable temperature regulation. This method overcomes the limitations of traditional fixed temperature range control, realizes dynamic adjustment of the target temperature according to the current fermentation state, and significantly improves the accuracy of temperature control. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 The present invention is a schematic flow chart of a method for controlling temperature in a beer fermentation process. DETAILED DESCRIPTION

[0016] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. Example

[0017] See also Figure 1 In this embodiment, a method for controlling the temperature of a beer fermentation process includes: S1. Establishing a fermentation knowledge base, wherein the fermentation knowledge base includes beer types and corresponding fermentation knowledge; the fermentation knowledge includes fermentation stages, target temperature acquisition models corresponding to the fermentation stages, and stage information; the stage information includes stage duration and stage characteristics; The fermentation stages include, in order: an adaptation stage, a logarithmic growth stage, a stable stage, and a decay stage; the stage characteristics are represented by a characteristic parameter array under each stage, and the characteristic parameter array includes three elements: an average carbon dioxide production rate, an average specific gravity change rate, and an average pH change rate; the stage duration is the maximum duration under each stage; The fermentation state parameters include carbon dioxide production rate, specific gravity change rate, pH change rate, and concentrations of alcohols, esters, and diacetyl.

[0018] In this embodiment, a knowledge base containing beer fermentation knowledge is established. Since different types of beer have different fermentation requirements, fermentation knowledge is established for each type of beer. This fermentation knowledge can be obtained through various channels such as professional books and practical experience; During the fermentation process of beer, the activity of yeast is not static, but presents different stage characteristics. Based on this characteristic, the fermentation process is divided into four stages: adaptation stage, logarithmic growth stage, stable stage and decline stage. This staged approach can more accurately describe and control the fermentation process, thereby improving the quality and flavor consistency of beer.

[0019] It should be noted that the adaptation stage is the process of yeast adapting to the new environment. In this stage, yeast cells begin to absorb oxygen and nutrients to prepare for subsequent rapid growth. The logarithmic growth stage is a period of rapid reproduction of yeast cells. At this time, the metabolic activity of yeast is the most vigorous, and a large amount of fermentation by-products are produced. After entering the stable stage, the growth rate of yeast tends to be flat, and the main process is the fermentation of sugars to produce ethanol and carbon dioxide. The final decline stage is the late fermentation stage, when most of the fermentable sugars have been consumed, the yeast begins to autolyze, and at the same time, some flavor substances are converted.

[0020] It is worth noting that the fermentation state parameters in each stage usually present different characteristics. These characteristics not only reflect the progress of the fermentation process, but also provide an important basis for judging the current fermentation stage and controlling the fermentation process. Based on this, a characteristic parameter array can be set for each stage to represent its characteristics, wherein the element value in the characteristic parameter array of each stage can be obtained after average calculation based on the carbon dioxide production rate, specific gravity change rate and pH change rate recorded for multiple times of the same beer at each stage. By accurately grasping the characteristics of each stage, each stage can be better identified, and the temperature of the fermentation process can be more accurately controlled.

[0021] S2. Obtain the type of beer to be fermented, and obtain corresponding fermentation knowledge according to the type of beer to be fermented; In this embodiment, the type of beer to be fermented can be input by a staff member, and the fermentation knowledge is obtained from a fermentation knowledge base according to the corresponding relationship.

[0022] S3, set a timer, and start timing when fermentation starts, and collect fermentation state parameters and fermentation liquid temperature at the same time; The steps of collecting fermentation state parameters include: At preset intervals :Measure the volume of carbon dioxide discharged by flow meter and calculate the carbon dioxide production rate; measure the specific gravity of fermentation liquid by hydrometer and calculate the specific gravity change rate; measure the pH value of fermentation liquid by pH meter and calculate the pH change rate; At preset intervals , obtain fermentation broth samples and determine the concentrations of alcohols, esters and diacetyl using corresponding equipment.

[0023] In this embodiment, in order to accurately monitor the fermentation process, it is necessary to regularly collect fermentation state parameters and fermentation liquid temperature. These data not only reflect the real-time state of fermentation, but also provide an important basis for subsequent temperature control.

[0024] At the beginning of fermentation, the timer is set and started. The timer setting ensures that all data collection has a clear time stamp, which is convenient for subsequent data analysis and processing. At the same time, the fermentation status parameters and fermentation liquid temperature are collected regularly.

[0025] The collection of fermentation state parameters is divided into two frequencies: high-frequency collection and low-frequency collection. High-frequency collection is mainly for parameters that change rapidly, including carbon dioxide production rate, specific gravity change rate and pH change rate. For example, high-frequency data collection is performed every 10 minutes. Specifically, the volume of carbon dioxide discharged is measured by a flow meter, and the carbon dioxide generation rate is calculated based on the time interval; the specific gravity of the fermentation liquid is measured by a hydrometer, and the specific gravity change rate is calculated; the pH value of the fermentation liquid is measured by a pH meter, and the pH change rate is calculated. The changes in these parameters can timely reflect the dynamic characteristics of the fermentation process.

[0026] Low-frequency acquisition focuses on slower-changing but equally important parameters, including the concentrations of alcohols, esters, and diacetyl. For example, every 6 hours, a low-frequency collection is performed. This process requires obtaining a fermentation broth sample and then measuring the concentration of these substances through the corresponding analytical equipment. These parameters reflect the formation of flavor substances during the fermentation process and are crucial for controlling the final flavor of beer.

[0027] This combination of high and low frequency acquisition can not only capture the rapid changes in the fermentation process in a timely manner, but also comprehensively monitor the generation of fermentation products. This method not only ensures the timeliness of the data, but also avoids interference with the fermentation process caused by too frequent sampling. At the same time, all collected data are associated with the timestamp of the timer, providing reliable time series data for subsequent data analysis and fermentation stage judgment.

[0028] S4, determining the current fermentation stage according to the fermentation state parameters and stage characteristics collected recently; The step of determining the current fermentation stage according to the most recently collected fermentation state parameters and stage characteristics comprises: Check whether there is a determined current fermentation stage, if so, record the determined current fermentation stage as the first fermentation stage, otherwise record the adaptation stage as the first fermentation stage; Check the first fermentation stage for subsequent stages; If not present, the current fermentation stage is determined as the first fermentation stage; If present, the subsequent stage was obtained and recorded as the second fermentation stage; Obtaining stage features corresponding to the first fermentation stage and the second fermentation stage, and respectively calculating a first similarity and a second similarity according to the most recently collected fermentation state parameters; Obtain the duration of the current stage and the stage duration corresponding to the first fermentation stage to obtain a first weight and a second weight; Compare whether the first similarity multiplied by the first weight is greater than or equal to the second similarity multiplied by the second weight; if so, determine the current fermentation stage as the first fermentation stage; if not, determine the current fermentation stage as the second fermentation stage.

[0029] The first similarity is calculated by the following formula: , In the formula, represents the first similarity, represents the average carbon dioxide production rate in the stage characteristics corresponding to the first fermentation stage, Indicates the carbon dioxide production rate in the most recently collected fermentation state parameters. It represents the average specific gravity change rate in the stage characteristics corresponding to the first fermentation stage, Indicates the rate of change of specific gravity in the most recently collected fermentation state parameters. represents the average pH change rate in the stage characteristics corresponding to the first fermentation stage, Indicates the pH change rate of the most recently collected fermentation state parameters; The second similarity is calculated by the following formula: , In the formula, represents the second similarity, represents the average carbon dioxide production rate in the stage characteristics corresponding to the second fermentation stage, represents the average specific gravity change rate in the stage characteristics corresponding to the second fermentation stage, represents the average pH change rate in the stage characteristics corresponding to the second fermentation stage; The first weight is expressed by the following formula: , to calculate, where represents the first weight, Indicates the duration of the first fermentation stage, Indicates the duration of the first fermentation stage. represents the preset scaling factor and ; The second weight is calculated by the following formula: , to calculate, where Represents the second weight.

[0030] In this embodiment, since the fermentation state is different at different stages, it is crucial to determine the fermentation stage for accurately controlling the fermentation process; the three parameters of carbon dioxide production rate, specific gravity change rate and pH change rate can effectively reflect yeast activity and fermentation progress: the carbon dioxide production rate directly reflects the metabolic activity of yeast; the specific gravity change rate reflects the consumption rate of fermentable sugars; and the pH change rate reflects the production of organic acids. These parameters show different characteristics at different stages of fermentation, so they can be used as reliable indicators for determining the stage.

[0031] However, relying solely on these parameters may lead to misjudgment. Therefore, the time factor is introduced as an auxiliary basis for judgment. Each fermentation stage has an estimated duration, and this information is stored in the previously established knowledge base. By combining the similarity of the current state parameters and the time factor, the current fermentation stage can be judged more accurately.

[0032] The specific judgment process is as follows: First, check whether there is a determined current fermentation stage. If so, record it as the first fermentation stage; if not, record the adaptation stage as the first fermentation stage (usually corresponding to the first judgment, when there is no determined current fermentation stage). Then, check whether there is a subsequent stage in the first fermentation stage. If not, directly determine the current stage as the first fermentation stage (usually it has been determined as the last stage, i.e., the decline stage, in the previous judgment, so there is no subsequent stage, and all subsequent stages are decline stages). If there is a subsequent stage, record it as the second fermentation stage. If in the previous judgment, the current stage is determined to be the logarithmic growth stage, then the logarithmic growth stage is taken as the first fermentation stage, and the next stage of the logarithmic growth stage, i.e., the stable stage, is taken as the second fermentation stage, and then the current stage is determined based on the collected data and time.

[0033] Next, the similarity between the current state parameters and the characteristics of the first and second fermentation stages is calculated respectively. The calculation of similarity takes into account three factors: carbon dioxide production rate, specific gravity change rate, and pH change rate. At the same time, the weight is calculated based on the duration of the current stage and the corresponding stage duration. This weight reflects the influence of the time factor. The closer the duration of the first fermentation stage is to the corresponding stage duration of the first stage, the smaller the possibility that the current stage is still the first stage, and the smaller the weight. The greater the possibility of the corresponding second stage, the larger the weight.

[0034] Finally, the weighted similarities are compared. If the weighted similarity of the first fermentation stage is greater than or equal to the weighted similarity of the second fermentation stage, the current stage is determined to be the first fermentation stage; otherwise, it is determined to be the second fermentation stage. This judgment method combining the similarity of state parameters and time factors can effectively balance the immediate state and the overall trend, avoid misjudgment caused by instantaneous fluctuations, and thus more accurately determine the current fermentation stage.

[0035] S5, obtaining a corresponding target temperature acquisition model according to the determined current fermentation stage and recording it as the current model; inputting the fermentation state parameters collected most recently into the current model to obtain the target temperature; The process of establishing the target temperature acquisition model corresponding to the fermentation stage includes: According to the historical fermentation state parameters and the corresponding fermentation liquid temperature, the learning data set corresponding to each fermentation stage is obtained; Divide the corresponding learning data set into a training set and a validation set according to a preset ratio; Build a machine learning framework and define the input layer and output layer, set the loss function and optimizer, and build a machine learning model; The fermentation state parameters are used as input layer data, and the corresponding fermentation liquid temperature is used as output layer data. The training set is used to train the constructed machine learning model according to the set batch size and number of training rounds. The trained machine learning model is tested using the test set, and a machine learning model that meets the preset accuracy is output as the target temperature acquisition model corresponding to the fermentation stage.

[0036] The step of obtaining a learning data set corresponding to each fermentation stage according to the historical fermentation state parameters and the corresponding fermentation liquid temperature comprises: Based on the preset standards, batches whose fermentation results meet the preset standards are selected from the historical fermentation records; For each selected batch, for each fermentation stage in the fermentation process: obtain the fermentation state parameters in the fermentation process at that stage, and obtain the fermentation liquid temperature at the time point corresponding to the fermentation state parameters; and combine the obtained fermentation state parameters and fermentation liquid temperature into a piece of learning data for that stage; Multiple learning data of each fermentation stage are aggregated to form a learning data set corresponding to each fermentation stage.

[0037] In this embodiment, a target temperature acquisition model is established and used. The purpose of the target temperature acquisition model is to determine a suitable target temperature based on the current fermentation state parameters, so as to achieve precise temperature control. It should be noted that the relationship between the fermentation state parameters and the suitable temperature is very complex. For example, higher temperatures usually speed up the fermentation rate, resulting in an increase in the rate of carbon dioxide production, but it may also affect the metabolic pathway of yeast and change the production of flavor substances, namely alcohols, esters and diacetyl. In addition, the sensitivity to temperature is different at different stages. For example, in the adaptation stage, a slightly higher temperature may help yeast to quickly adapt to the environment, while in the later stages, a lower temperature may contribute to the formation of flavor. This complex nonlinear relationship is difficult to accurately describe with a simple mathematical formula, so a machine learning method is chosen to establish a model. The specific process is as follows: First, batches whose fermentation results meet the preset standards are selected from the historical fermentation records. It should be noted that the preset standards can be expert evaluation results, physical and chemical indicators, etc. For each selected batch, the fermentation state parameters and the fermentation liquid temperature at the corresponding time point are extracted according to the fermentation stage. These paired data constitute the learning data set for each stage. Next, the learning data set for each stage is divided into a training set and a validation set in a ratio of 8 to 2. The training set is used for model learning, and the validation set is used to evaluate the performance of the model.

[0038] Then, a machine learning framework is built. In this embodiment, a multi-layer perceptron (MLP) neural network is selected as the model structure. Specifically, a neural network containing two hidden layers is constructed, each hidden layer has 64 neurons, and ReLU is used as the activation function. The input layer contains 6 neurons, corresponding to 6 fermentation state parameters (carbon dioxide production rate, specific gravity change rate, pH change rate, alcohol concentration, ester concentration and diacetyl concentration), and the output layer is 1 neuron, corresponding to the target temperature. The mean square error (MSE) is selected as the loss function, and the Adam optimizer is used for model optimization.

[0039] During the training process, the batch size is set to 32 and the number of training rounds is set to 1000. For example, assuming there are 10,000 training data, the model will be updated 312 times in each round of training (10000 / 32≈312). After 1000 rounds of training, the model will be updated about 312,000 times in total. During this process, the model continuously adjusts its internal parameters to minimize the difference between the predicted temperature and the actual temperature.

[0040] After training is completed, the model is tested using the validation set. If the mean absolute error of the model on the validation set is less than 0.3°C, the model is considered to have reached the preset accuracy standard and can be used as the target temperature acquisition model for this fermentation stage.

[0041] In this embodiment, once the current fermentation stage is determined, the corresponding target temperature acquisition model can be called. The recently collected fermentation state parameters are input into this model, and the model will output a target temperature. This target temperature is the appropriate temperature calculated by the machine learning algorithm based on historical data and current state.

[0042] This method can capture the complex nonlinear relationship in the fermentation process and is more accurate than the traditional fixed temperature control method. As new fermentation data continues to accumulate, the model can be retrained regularly to continuously optimize and adapt to new situations. In this way, the most suitable temperature can be obtained at each fermentation stage, providing precise guidance for subsequent temperature control, thereby improving the controllability of the beer fermentation process and the quality stability of the final product.

[0043] S6. Regulate according to the fermentation liquid temperature and the target temperature.

[0044] The regulation according to the fermentation liquid temperature and the target temperature includes: Real-time monitoring of fermentation liquid temperature and target temperature; When the deviation between the fermentation liquid temperature and the target temperature exceeds a preset threshold, the power of the temperature control device is adjusted accordingly until the fermentation liquid temperature reaches the target temperature; and when the fermentation liquid temperature reaches the target temperature value, the temperature control device is set to the insulation mode.

[0045] In this embodiment, the fermentation liquid temperature and the target temperature are monitored; when it is detected that the deviation between the actual temperature of the fermentation liquid and the target temperature exceeds a preset threshold value such as ±0.5, the system will automatically adjust the power of the temperature control device. Specifically, if the actual temperature is higher than the target temperature, the system will increase the cooling power or reduce the heating power; if the actual temperature is lower than the target temperature, the system will increase the heating power or reduce the cooling power. The temperature control system can use a PID (proportional-integral-differential) control algorithm to dynamically adjust the power output of the temperature control device according to the size, duration and change trend of the temperature deviation. This method can achieve fast response and stable temperature regulation, and control the temperature fluctuation within the minimum range. When the fermentation liquid temperature reaches the target temperature, the system will automatically switch the temperature control device to the insulation mode to maintain temperature stability with minimal energy consumption.

[0046] In the several embodiments provided by the present invention, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of the units is only one, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0047] The above description is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

[0048] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for controlling the temperature of a beer fermentation process, characterized in that: include: Establishing a fermentation knowledge base, the fermentation knowledge base includes beer types and corresponding fermentation knowledge; the fermentation knowledge includes fermentation stages, target temperature acquisition models corresponding to the fermentation stages, and stage information; the stage information includes stage duration and stage characteristics; Obtain the type of beer to be fermented, and obtain corresponding fermentation knowledge according to the type of beer to be fermented; Set a timer and start timing when fermentation starts, while collecting fermentation state parameters and fermentation liquid temperature; Determine the current fermentation stage according to the fermentation state parameters and stage characteristics collected recently; According to the determined current fermentation stage, a corresponding target temperature acquisition model is obtained and recorded as the current model; the most recently collected fermentation state parameters are input into the current model to obtain the target temperature; Regulate according to the fermentation liquid temperature and target temperature.

2. A beer fermentation process temperature control method according to claim 1, characterized in that: The fermentation stages include, in order: an adaptation stage, a logarithmic growth stage, a stable stage, and a decay stage; the stage characteristics are represented by a characteristic parameter array under each stage, and the characteristic parameter array includes three elements: an average carbon dioxide production rate, an average specific gravity change rate, and an average pH change rate; the stage duration is the maximum duration under each stage; The fermentation state parameters include carbon dioxide production rate, specific gravity change rate, pH change rate, and concentrations of alcohols, esters, and diacetyl.

3. A beer fermentation process temperature control method according to claim 2, characterized in that: The step of collecting fermentation state parameters comprises: At preset intervals :Measure the volume of carbon dioxide discharged by flow meter and calculate the carbon dioxide production rate; measure the specific gravity of fermentation liquid by hydrometer and calculate the specific gravity change rate; measure the pH value of fermentation liquid by pH meter and calculate the pH change rate; At preset intervals , obtain fermentation broth samples and determine the concentrations of alcohols, esters and diacetyl using corresponding equipment.

4. A beer fermentation process temperature control method according to claim 3, characterized in that: The step of determining the current fermentation stage according to the most recently collected fermentation state parameters and stage characteristics comprises: Check whether there is a determined current fermentation stage, if so, record the determined current fermentation stage as the first fermentation stage, otherwise record the adaptation stage as the first fermentation stage; Check the first fermentation stage for subsequent stages; If not present, the current fermentation stage is determined as the first fermentation stage; If present, the subsequent stage was obtained and recorded as the second fermentation stage; Obtaining stage features corresponding to the first fermentation stage and the second fermentation stage, and respectively calculating a first similarity and a second similarity according to the most recently collected fermentation state parameters; Obtain the duration of the current stage and the stage duration corresponding to the first fermentation stage to obtain a first weight and a second weight; Compare whether the first similarity multiplied by the first weight is greater than or equal to the second similarity multiplied by the second weight; if so, determine the current fermentation stage as the first fermentation stage; if not, determine the current fermentation stage as the second fermentation stage.

5. A beer fermentation process temperature control method according to claim 4, characterized in that: The first similarity is calculated by the following formula: ； In the formula, represents the first similarity, represents the average carbon dioxide production rate in the stage characteristics corresponding to the first fermentation stage, Indicates the carbon dioxide production rate in the most recently collected fermentation state parameters. It represents the average specific gravity change rate in the stage characteristics corresponding to the first fermentation stage, Indicates the rate of change of specific gravity in the most recently collected fermentation state parameters. represents the average pH change rate in the stage characteristics corresponding to the first fermentation stage, Indicates the pH change rate of the most recently collected fermentation state parameters; The second similarity is calculated by the following formula: ； In the formula, represents the second similarity, represents the average carbon dioxide production rate in the stage characteristics corresponding to the second fermentation stage, represents the average specific gravity change rate in the stage characteristics corresponding to the second fermentation stage, represents the average pH change rate in the stage characteristics corresponding to the second fermentation stage; The first weight is expressed by the following formula: , to calculate, where represents the first weight, Indicates the duration of the first fermentation stage, Indicates the duration of the first fermentation stage. represents the preset scaling factor and ; The second weight is calculated by the following formula: , to calculate, where Represents the second weight.

6. A beer fermentation process temperature control method according to claim 1, characterized in that: The process of establishing the target temperature acquisition model corresponding to the fermentation stage includes: According to the historical fermentation state parameters and the corresponding fermentation liquid temperature, the learning data set corresponding to each fermentation stage is obtained; Divide the corresponding learning data set into a training set and a validation set according to a preset ratio; Build a machine learning framework and define the input layer and output layer, set the loss function and optimizer, and build a machine learning model; The fermentation state parameters are used as input layer data, and the corresponding fermentation liquid temperature is used as output layer data. The training set is used to train the constructed machine learning model according to the set batch size and number of training rounds. The trained machine learning model is tested using the test set, and a machine learning model that meets the preset accuracy is output as the target temperature acquisition model corresponding to the fermentation stage.

7. A beer fermentation process temperature control method according to claim 6, characterized in that: The step of obtaining a learning data set corresponding to each fermentation stage according to the historical fermentation state parameters and the corresponding fermentation liquid temperature comprises: Based on the preset standards, batches whose fermentation results meet the preset standards are selected from the historical fermentation records; For each selected batch, for each fermentation stage in the fermentation process: obtain the fermentation state parameters in the fermentation process at this stage, obtain the fermentation liquid temperature at the time point corresponding to the fermentation state parameters; and form a learning data for this stage with the obtained fermentation state parameters and fermentation liquid temperature; Multiple learning data of each fermentation stage are aggregated to form a learning data set corresponding to each fermentation stage.

8. A beer fermentation process temperature control method according to claim 7, characterized in that: The regulation according to the fermentation liquid temperature and the target temperature includes: Real-time monitoring of fermentation liquid temperature and target temperature; When the deviation between the fermentation liquid temperature and the target temperature exceeds a preset threshold, the power of the temperature control device is adjusted accordingly until the fermentation liquid temperature reaches the target temperature; and when the fermentation liquid temperature reaches the target temperature value, the temperature control device is set to the insulation mode.

Images