White spirit brewing process parameter optimization method based on artificial intelligence

Abstract

The invention relates to the technical field of food processing parameter optimization, and particularly discloses a Baijiu brewing process parameter optimization method based on artificial intelligence, which comprises the following steps of: acquiring temperature and steam flow data in a distillation process in real time, performing discrete Fourier transform processing on the temperature and steam flow data, and calculating a temperature fluctuation index and a steam flow fluctuation index; the two indexes are used as input characteristics, a wine quality prediction model is established by using a deep neural network technology, the model can predict whether the wine quality is qualified or not according to real-time data, and for the wine with an unqualified prediction result, the system can automatically analyze reasons, so that the wine quality prediction accuracy is improved. According to the method, through real-time monitoring and intelligent adjustment, the product quality problem caused by parameter fluctuation in a traditional brewing process is effectively solved, the production efficiency is improved, and the stability of the white spirit quality is ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of food processing parameter optimization, and in particular to a method for optimizing liquor brewing process parameters based on artificial intelligence. Background Art

[0002] As part of traditional Chinese culture, liquor has a long history and unique brewing process. The brewing of traditional liquor mainly relies on the experience accumulation and technical inheritance of masters, among which distillation is a crucial part of the liquor brewing process; the control of temperature and steam flow in the distillation process directly affects the flavor, aroma and final quality of the liquor; however, in actual production, due to changes in the external environment, aging of equipment and uncertainty in manual operation, it is difficult to always maintain ideal distillation conditions, resulting in fluctuations in liquor quality.

[0003] In recent years, with the development of Internet of Things technology, big data analysis and artificial intelligence technology, industrial production is gradually moving towards intelligence. The application of these emerging technologies has brought new opportunities to the traditional brewing industry. Sensors can monitor various parameters in the distillation process in real time, big data platforms can store and process large amounts of production data, and artificial intelligence can mine valuable patterns from them to help decision makers better understand and control the production process.

[0004] At present, there are still few cases of applying these advanced technologies in the liquor industry, especially in the optimization of process parameters. Existing control systems can only provide static data records, lack dynamic adjustment functions, and cannot provide real-time warnings and corrections for abnormal situations during the production process. In addition, due to the lack of sufficient data support, existing prediction models are also limited in accuracy and practicality. Therefore, how to use modern information technology to monitor the key parameters in the liquor brewing process in real time and dynamically adjust the process conditions based on these data to ensure the consistency and stability of product quality has become an urgent problem to be solved. Summary of the invention

[0005] The purpose of the present invention is to provide a method for optimizing liquor brewing process parameters based on artificial intelligence to solve the problems in the above background.

[0006] The purpose of the present invention can be achieved through the following technical solutions: A method for optimizing liquor brewing process parameters based on artificial intelligence, comprising the following steps: S1: Obtain the temperature data of liquor during the distillation process, analyze the temperature data, calculate the temperature fluctuation index, and evaluate the temperature stability during the distillation process; S2: Obtain the steam flow data of liquor during the distillation process, analyze the steam flow data, calculate the steam flow fluctuation index, and determine whether the concentration of liquor during the distillation process is normal; S3: Comprehensively analyze the temperature fluctuation index and the steam flow fluctuation index, extract the temperature fluctuation index characteristics and the steam flow fluctuation index characteristics, establish a wine quality prediction model, and predict whether the wine quality is qualified according to the prediction model; S4: Based on the prediction results, the wine quality is divided into qualified wine and unqualified wine, the temperature data and steam flow data of the unqualified wine are adjusted in time, and the wine quality prediction is re-performed.

[0007] As a further solution of the present invention: the step of obtaining the temperature data of liquor during the distillation process, analyzing the temperature data, and calculating the temperature fluctuation index specifically includes: The temperature data of the distillation process is collected in time series, and the collected temperature data is subjected to discrete Fourier transform. The calculation expression is: ; Where k represents the number of frequency components of the Fourier transform result, represents the kth frequency component, n represents the number of data points, j represents the number of temperature data, represents the jth temperature data point in the time series; Calculate the amplitude spectrum, the calculation expression is: ; In the formula, represents the real part of the Fourier transform result, The imaginary part of the Fourier transform result; The temperature fluctuation index is calculated using the amplitude spectrum. The calculation expression is: ; In the formula, Indicates the temperature fluctuation index.

[0008] As a further solution of the present invention: the calculation of the temperature fluctuation index and the evaluation of the temperature stability during the distillation process specifically include: It is determined whether the temperature fluctuation index is greater than or equal to a preset temperature fluctuation index threshold. If so, the corresponding distillation temperature is unstable; if not, the corresponding distillation temperature is stable.

[0009] As a further solution of the present invention: the obtaining of steam flow data of liquor during distillation, analyzing the steam flow data, and calculating the steam flow fluctuation index specifically include: Collect steam flow data during the distillation process in time series; Calculate the sum of squares of the deviations between the flow rate at each time point and the average steam flow rate. The calculation expression is: ; In the formula, i represents the number of collected steam flow data, represents the sum of squared deviations of all steam flow data, represents the i-th steam flow data, Represents the mean of all steam flow data; Calculate the steam flow fluctuation index, the calculation expression is: ; In the formula, represents the steam flow fluctuation index, and m represents the total number of steam flow data.

[0010] As a further solution of the present invention: the calculation of the steam flow fluctuation index to determine whether the concentration of the liquor during the distillation process is normal specifically includes: Determine the steam flow fluctuation index Is it greater than or equal to the preset steam flow fluctuation index threshold? If so, the alcohol concentration in the distillation process is abnormal; if not, the alcohol concentration in the distillation process is normal.

[0011] As a further solution of the present invention: the comprehensive analysis of the temperature fluctuation index and the steam flow fluctuation index, the extraction of the temperature fluctuation index characteristics and the steam flow fluctuation index characteristics, and the establishment of the wine quality prediction model specifically include: The temperature fluctuation index and the steam flow fluctuation index are multiplied and calculated, and recorded as the correlation index; The temperature fluctuation index and the steam flow fluctuation index are calculated by difference and the absolute value is taken, which is recorded as the deviation index; Extract temperature fluctuation index features, steam flow fluctuation index features, correlation index features and deviation index features as inputs of the model; Select a deep neural network as the model and build a model to predict wine quality; The output of the model is the probability that the wine quality is qualified, and the calculation expression is: ; In the formula, represents the input feature vector, It indicates the probability that the wine quality is qualified. represents the mass fraction of liquor, represents the weight matrix, represents the bias term, represents the activation function; The model is trained using accurate temperature fluctuation index, steam flow fluctuation index, correlation index and deviation index in the historical distillation process as a set of historical data; The cross entropy loss function is used for training, and the calculation expression is: ; Where L represents the loss function, a represents the number of historical training data sets, a=1,2,...,N, represents the mass fraction of the ath liquor, The feature vector representing the a-th group of historical data; Use the gradient descent optimization algorithm to update the model parameters W and b to minimize the loss function; Real-time data including real-time temperature fluctuation index, real-time steam flow fluctuation index, real-time correlation index and real-time deviation index are obtained as test data sets to evaluate model performance.

[0012] As a further solution of the present invention: the predicting whether the wine quality is qualified according to the prediction model specifically includes: During the distillation process, the real-time temperature fluctuation index, steam flow fluctuation index, correlation index and deviation index are input into the model, and the liquor quality is predicted. The model will output the predicted probability of the liquor quality. The probability of the wine quality being qualified is compared with the preset probability threshold of the wine quality being qualified. If the probability of the wine quality being qualified is greater than or equal to the preset probability threshold of the wine quality being qualified, the corresponding wine quality is qualified, and the corresponding wine quality Q output value is 1. If the probability of the wine quality being qualified is less than the preset probability threshold of the wine quality being qualified, the corresponding wine quality is unqualified, and the corresponding wine quality Q output value is 0.

[0013] As a further solution of the present invention: the wine quality is divided into qualified wine and unqualified wine according to the prediction result, the temperature data and steam flow data of the unqualified wine are adjusted in time, and the wine quality prediction is re-performed, which specifically includes: Determine whether the probability of the wine quality being qualified is greater than or equal to a preset probability threshold of the wine quality being qualified. If so, record it as qualified wine; if not, record it as unqualified wine; Obtaining the temperature fluctuation index of the predicted unqualified liquor, and comparing the temperature fluctuation index with the corresponding temperature fluctuation index threshold; if the temperature fluctuation index is greater than or equal to the temperature fluctuation index threshold, the temperature data in the distillation process causes the quality of the liquor to be unqualified, and the temperature data is adjusted; Obtaining the steam flow fluctuation index of the predicted unqualified liquor, and comparing the steam flow fluctuation index with the corresponding steam flow fluctuation index threshold; if the steam flow fluctuation index is greater than or equal to the steam flow fluctuation index threshold, the steam flow data in the distillation process causes the quality of the liquor to be unqualified, and the steam flow data is adjusted; After adjusting the temperature data and steam flow data, the wine quality continues to be monitored and evaluated according to the time series.

[0014] Beneficial effects of the present invention: (1) The present invention uses artificial intelligence technology to monitor and analyze the key parameters of temperature and steam flow data in the distillation process of liquor in real time, thereby significantly improving the stability of liquor quality. Traditional liquor brewing relies on the manual operation and technical accumulation of experienced winemakers, but in large-scale industrial production, it is difficult to ensure that each distillation can reach the optimal state, especially when facing environmental changes or equipment aging, which can easily cause fluctuations in liquor quality. The present invention establishes a quality prediction model based on deep learning, which can accurately capture any tiny fluctuations in the distillation process and quantify the impact of these changes on liquor quality by calculating the temperature fluctuation index and steam flow fluctuation index. Once a trend of deviation from the standard is found, the system will automatically trigger the adjustment mechanism to quickly correct the process parameters and return the distillation conditions to the ideal state. This instant feedback and adjustment capability enables the consistency of liquor quality to be maintained even on a continuous production line, greatly reducing product quality problems caused by human factors or environmental interference, thereby improving the controllability and reliability of the entire brewing process. (2) Under the technical framework of the present invention, the temperature fluctuation index and the steam flow fluctuation index are not only used as indicators for evaluating the stability of the distillation process, but also become an important basis for adjusting the process conditions in real time; once it is detected that the quality of a batch of liquor does not meet expectations, the system can automatically or assist the winemaker to adjust the temperature and steam flow according to the pre-set threshold until they are restored to the optimal working state; this instant feedback mechanism greatly simplifies the workload of the winemaker, allowing even less experienced operators to produce high-quality liquor; In addition, through continuous data collection and model training, the present invention can also gradually improve the accuracy of the prediction model, making it more adaptable to different types of liquor production and complex and changeable actual production environments; as time goes by, the system will accumulate more and more experience, further optimize the selection of process parameters, reduce trial and error costs, and improve overall production efficiency; in summary, the present invention not only solves the problem of imprecise adjustment of process parameters in traditional brewing methods, but also brings a new opportunity for the liquor industry to transform from experience-driven to data-driven, and promotes industrial technology upgrading. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The present invention will be further described below in conjunction with the accompanying drawings.

[0016] Figure 1 It is a flowchart of the specific steps of a method for optimizing liquor brewing process parameters based on artificial intelligence of the present invention. DETAILED DESCRIPTION

[0017] 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.

[0018] See also Figure 1 As shown, the present invention is a method for optimizing liquor brewing process parameters based on artificial intelligence, comprising the following steps: S1: Obtain the temperature data of liquor during the distillation process, analyze the temperature data, calculate the temperature fluctuation index, and evaluate the temperature stability during the distillation process; S2: Obtain the steam flow data of liquor during the distillation process, analyze the steam flow data, calculate the steam flow fluctuation index, and determine whether the concentration of liquor during the distillation process is normal; S3: Comprehensively analyze the temperature fluctuation index and the steam flow fluctuation index, extract the temperature fluctuation index characteristics and the steam flow fluctuation index characteristics, establish a wine quality prediction model, and predict whether the wine quality is qualified according to the prediction model; S4: Based on the prediction results, the wine quality is divided into qualified wine and unqualified wine, the temperature data and steam flow data of the unqualified wine are adjusted in time, and the wine quality prediction is re-performed.

[0019] In S1, the temperature data of liquor during the distillation process is obtained, the temperature data is analyzed, and the temperature stability during the distillation process is evaluated, including: The temperature data of the distillation process is collected in time series (such as every minute), and the collected temperature data is subjected to discrete Fourier transform. The calculation expression is: ; Where k represents the number of frequency components of the Fourier transform result, represents the kth frequency component, n represents the number of data points, j represents the number of temperature data, represents the jth temperature data point in the time series; Calculate the amplitude spectrum, the calculation expression is: ; In the formula, represents the real part of the Fourier transform result, The imaginary part of the Fourier transform result; The temperature fluctuation index is calculated using the amplitude spectrum. The calculation expression is: ; In the formula, Indicates the temperature fluctuation index; It should be noted that the temperature fluctuation index reflects the temperature change of liquor during the distillation process, and the larger the temperature fluctuation index, the more unstable the corresponding distillation temperature; Temperature Fluctuation Index Compare with preset temperature fluctuation index threshold; If the temperature fluctuation index If it is greater than or equal to the preset temperature fluctuation index threshold, it means that the corresponding distillation temperature is unstable; If the temperature fluctuation index If it is less than the preset temperature fluctuation index threshold, it means that the corresponding distillation temperature is stable.

[0020] In S2, the steam flow data of the liquor during the distillation process is obtained, the steam flow data is analyzed, the steam flow fluctuation index is calculated, and whether the extraction efficiency of the liquor during the distillation process is normal is determined, which specifically includes: Collect steam flow data during the distillation process in a time series (e.g., every minute) and calculate the average of the collected steam flow data ; Calculate the sum of squares of the deviations between the flow rate at each time point and the average steam flow rate. The calculation expression is: ; Where i represents the number of collected steam flow data, i=1,2,...,m, represents the sum of squared deviations of all steam flow data, represents the i-th steam flow data, Represents the mean of all steam flow data; Calculate the steam flow fluctuation index, the calculation expression is: ; In the formula, represents the steam flow fluctuation index, and m represents the total number of steam flow data; It should be noted that: Steam flow fluctuation index The larger the value, the more abnormal the concentration of the corresponding liquor will be during the distillation process, which will lead to worse quality of the liquor. Steam flow fluctuation index Compared with the preset steam flow fluctuation index threshold, if the steam flow fluctuation index If the steam flow fluctuation index is greater than or equal to the preset steam flow fluctuation index threshold, it means that the alcohol concentration in the corresponding distillation process is abnormal. If it is less than the preset steam flow fluctuation index threshold, it means that the alcohol concentration in the corresponding distillation process is normal.

[0021] In S3, the temperature fluctuation index and the steam flow fluctuation index are comprehensively analyzed, the temperature fluctuation index characteristics and the steam flow fluctuation index characteristics are extracted, and a wine quality prediction model is established. The prediction of whether the wine quality is qualified according to the prediction model specifically includes: The temperature fluctuation index and the steam flow fluctuation index are multiplied and calculated, and recorded as the correlation index; The temperature fluctuation index and the steam flow fluctuation index are calculated by difference and the absolute value is taken, which is recorded as the deviation index; Extract temperature fluctuation index features, steam flow fluctuation index features, correlation index features and deviation index features as inputs of the model; Input feature vector Including temperature fluctuation index characteristics, steam flow fluctuation index characteristics, correlation index characteristics and deviation index characteristics; Select a deep neural network as the model and build a model to predict the quality of the wine (qualified or unqualified); The output of the model is the probability that the wine quality is qualified, and the calculation expression is: ; In the formula, represents the input feature vector, It indicates the probability that the wine quality is qualified. represents the mass fraction of liquor, represents the weight matrix, represents the bias term, represents the activation function; The model is trained using accurate temperature fluctuation index, steam flow fluctuation index, correlation index and deviation index in the historical distillation process as a set of historical data; The cross entropy loss function is used for training, and the calculation expression is: ; Where L represents the loss function, a represents the number of historical training data sets, a=1,2,...,N, represents the mass fraction of the ath liquor, The feature vector representing the a-th group of historical data; Use the gradient descent optimization algorithm to update the model parameters W and b to minimize the loss function; Obtain real-time data including real-time temperature fluctuation index, real-time steam flow fluctuation index, real-time correlation index and real-time deviation index as test data sets to evaluate model performance, and calculate the model's accuracy index, precision index, recall index and F1 value index; During the distillation process, the real-time temperature fluctuation index, steam flow fluctuation index, correlation index and deviation index are input into the model, and the liquor quality is predicted. The model will output the predicted value of the liquor quality, specifically: The probability of the wine quality being qualified is compared with the preset probability threshold of the wine quality being qualified. If the probability of the wine quality being qualified is greater than or equal to the preset probability threshold of the wine quality being qualified, it means that the corresponding wine quality is qualified, and the corresponding wine quality Q output value is 1. If the probability of the wine quality being qualified is less than the preset probability threshold of the wine quality being qualified, it means that the corresponding wine quality is unqualified, and the corresponding wine quality Q output value is 0. It should be noted that the prediction model predicts whether the liquor quality is qualified. It can timely judge the liquor quality if there are abnormal process parameters in the liquor brewing process, so as to make corresponding adjustments.

[0022] In S4, according to the prediction results, the liquor quality is divided into qualified liquor and unqualified liquor, the temperature data and steam flow data of the unqualified liquor are adjusted in time, and the liquor quality prediction is re-performed, which specifically includes: Determine whether the probability of the wine quality being qualified is greater than or equal to a preset probability threshold of the wine quality being qualified. If so, record it as qualified wine; if not, record it as unqualified wine; Obtain the prediction results of unqualified liquor quality, and analyze whether the reason for the unqualified liquor quality is caused by the temperature data or the steam flow data during the distillation process; Obtain the temperature fluctuation index of the predicted unqualified liquor, and compare the temperature fluctuation index with the corresponding temperature fluctuation index threshold. If the temperature fluctuation index is greater than or equal to the temperature fluctuation index threshold, it means that the temperature data in the distillation process causes the quality of the liquor to be unqualified, and the temperature data needs to be adjusted to ensure that the temperature data is stable. Obtain the steam flow fluctuation index of the predicted unqualified liquor, and compare the steam flow fluctuation index with the corresponding steam flow fluctuation index threshold. If the steam flow fluctuation index is greater than or equal to the steam flow fluctuation index threshold, it means that the steam flow data in the distillation process causes the quality of the liquor to be unqualified, and the steam flow data needs to be adjusted to ensure that the steam flow data is stable. After adjusting the temperature data and steam flow data, it is necessary to continue monitoring and evaluating the wine quality according to the time series (such as every minute); It should be noted that during the wine distillation process, the temperature and steam flow may change at any time, resulting in the wine quality being unqualified. Timely prediction of the wine quality can effectively avoid the temperature and steam flow causing the wine quality to be unqualified.

[0023] The working principle of the present invention is to improve the distillation quality control level in the process of liquor production through intelligent means, ensure the excellent quality of the final product, obtain the temperature data of liquor in the distillation process, and analyze it. Through the discrete Fourier transform of these temperature data, the amplitude spectrum is calculated, and then the temperature fluctuation index is obtained to evaluate the stability of the temperature in the distillation process; the temperature fluctuation index reflects the change of temperature in the distillation process, and its value is inversely proportional to the stability of the distillation temperature. By comparing the temperature fluctuation index with the preset threshold, it can be judged whether the temperature is stable; the steam flow data in the distillation process of liquor is collected and analyzed, and the steam flow fluctuation index is calculated; by calculating the square sum of the deviations between the steam flow and the average flow at each time point, the steam flow fluctuation index is further obtained to judge whether the concentration of the liquor in the distillation process is normal; the steam flow fluctuation index is also compared with the preset threshold to determine whether the steam flow needs to be adjusted to ensure that the concentration of the liquor is within a reasonable range; the temperature fluctuation index and the steam flow fluctuation index are comprehensively analyzed, the features are extracted, and a liquor quality prediction model is established based on this. This model uses deep neural network technology to predict the probability of qualified wine quality by inputting feature vectors including temperature fluctuation index, steam flow fluctuation index, correlation index and deviation index; the model has been trained with historical data and can effectively identify the key factors affecting the quality of the wine, and make a qualified or unqualified judgment based on this; based on the model prediction results, the wine is divided into qualified and unqualified categories; for unqualified wine, the cause is analyzed, which may be caused by abnormal temperature or steam flow parameters; corresponding adjustment measures are taken for different reasons, such as adjusting the temperature or steam flow, until the quality is predicted again after meeting the standards, to ensure that the final product meets the quality requirements.

[0024] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0025] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center by wired or wireless (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0026] It should be understood that the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. A and B can be singular or plural. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship, but it may also indicate an "and / or" relationship. Please refer to the context for specific understanding.

[0027] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0028] The above is a detailed description of an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. A liquor brewing process parameter optimization method based on artificial intelligence, characterized in that: The following steps are involved: S1: Obtain the temperature data of liquor during the distillation process, analyze the temperature data, calculate the temperature fluctuation index, and evaluate the temperature stability during the distillation process; S2: Obtain the steam flow data of liquor during the distillation process, analyze the steam flow data, calculate the steam flow fluctuation index, and determine whether the concentration of liquor during the distillation process is normal; S3: Comprehensively analyze the temperature fluctuation index and the steam flow fluctuation index, extract the temperature fluctuation index characteristics and the steam flow fluctuation index characteristics, establish a wine quality prediction model, and predict whether the wine quality is qualified according to the prediction model; S4: Based on the prediction results, the wine quality is divided into qualified wine and unqualified wine, the temperature data and steam flow data of the unqualified wine are adjusted in time, and the wine quality prediction is re-performed.

2. According to the artificial intelligence-based liquor brewing process parameter optimization method of claim 1, it is characterized in that: The obtaining of temperature data of liquor during the distillation process, analyzing the temperature data, and calculating the temperature fluctuation index specifically include: The temperature data of the distillation process is collected in time series, and the collected temperature data is subjected to discrete Fourier transform. The calculation expression is: ; Where k represents the number of frequency components of the Fourier transform result, represents the kth frequency component, n represents the number of data points, j represents the number of temperature data, represents the jth temperature data point in the time series; Calculate the amplitude spectrum, the calculation expression is: ; In the formula, represents the real part of the Fourier transform result, The imaginary part of the Fourier transform result; The temperature fluctuation index is calculated using the amplitude spectrum. The calculation expression is: ; In the formula, Indicates the temperature fluctuation index.

3. According to claim 2, a method for optimizing liquor brewing process parameters based on artificial intelligence is characterized in that: The calculation of the temperature fluctuation index and the evaluation of the temperature stability during the distillation process specifically include: It is determined whether the temperature fluctuation index is greater than or equal to a preset temperature fluctuation index threshold. If so, the corresponding distillation temperature is unstable; if not, the corresponding distillation temperature is stable.

4. The method for optimizing liquor brewing process parameters based on artificial intelligence according to claim 1 is characterized in that: The method of obtaining steam flow data during the distillation of liquor, analyzing the steam flow data, and calculating the steam flow fluctuation index specifically includes: Collect steam flow data during the distillation process in time series; Calculate the sum of squares of the deviations between the flow rate at each time point and the average steam flow rate. The calculation expression is: ; In the formula, i represents the number of collected steam flow data, represents the sum of squared deviations of all steam flow data, represents the i-th steam flow data, Represents the mean of all steam flow data; calculate the steam flow fluctuation index, the calculation expression is: ; In the formula, represents the steam flow fluctuation index, and m represents the total number of steam flow data.

5. According to the artificial intelligence-based liquor brewing process parameter optimization method of claim 4, it is characterized in that: The calculation of the steam flow fluctuation index to determine whether the concentration of the liquor during the distillation process is normal specifically includes: Determine the steam flow fluctuation index Is it greater than or equal to the preset steam flow fluctuation index threshold? If so, the alcohol concentration in the distillation process is abnormal; if not, the alcohol concentration in the distillation process is normal.

6. The method for optimizing liquor brewing process parameters based on artificial intelligence according to claim 1 is characterized in that: The comprehensive analysis of the temperature fluctuation index and the steam flow fluctuation index, extraction of the temperature fluctuation index characteristics and the steam flow fluctuation index characteristics, and establishment of the wine quality prediction model specifically include: The temperature fluctuation index and the steam flow fluctuation index are multiplied and calculated, and recorded as the correlation index; The temperature fluctuation index and the steam flow fluctuation index are calculated by difference and the absolute value is taken, which is recorded as the deviation index; Extract temperature fluctuation index features, steam flow fluctuation index features, correlation index features and deviation index features as inputs of the model; Select a deep neural network as the model and build a model to predict wine quality; The output of the model is the probability that the wine quality is qualified, and the calculation expression is: ; In the formula, represents the input feature vector, It indicates the probability that the wine quality is qualified. represents the mass fraction of liquor, represents the weight matrix, represents the bias term, represents the activation function; The model is trained using accurate temperature fluctuation index, steam flow fluctuation index, correlation index and deviation index in the historical distillation process as a set of historical data; The cross entropy loss function is used for training, and the calculation expression is: ; Where L represents the loss function, a represents the number of historical training data sets, a=1,2,...,N, represents the mass fraction of the ath liquor, The feature vector representing the a-th group of historical data; Use the gradient descent optimization algorithm to update the model parameters W and b to minimize the loss function; Real-time data including real-time temperature fluctuation index, real-time steam flow fluctuation index, real-time correlation index and real-time deviation index are obtained as test data sets to evaluate model performance.

7. The method for optimizing liquor brewing process parameters based on artificial intelligence according to claim 1 is characterized in that: The step of predicting whether the wine quality is qualified according to the prediction model specifically includes: During the distillation process, the real-time temperature fluctuation index, steam flow fluctuation index, correlation index and deviation index are input into the model, and the liquor quality is predicted. The model will output the predicted probability of the liquor quality. The probability of the wine quality being qualified is compared with the preset probability threshold of the wine quality being qualified. If the probability of the wine quality being qualified is greater than or equal to the preset probability threshold of the wine quality being qualified, the corresponding wine quality is qualified, and the corresponding wine quality Q output value is 1. If the probability of the wine quality being qualified is less than the preset probability threshold of the wine quality being qualified, the corresponding wine quality is unqualified, and the corresponding wine quality Q output value is 0.

8. The method for optimizing liquor brewing process parameters based on artificial intelligence according to claim 1 is characterized in that: The wine quality is divided into qualified wine and unqualified wine according to the prediction result, the temperature data and steam flow data of the unqualified wine are adjusted in time, and the wine quality prediction is re-performed, which specifically includes: Determine whether the probability of the wine quality being qualified is greater than or equal to a preset probability threshold of the wine quality being qualified. If so, record it as qualified wine; if not, record it as unqualified wine; Obtain the temperature fluctuation index of the predicted unqualified liquor, and compare the temperature fluctuation index with the corresponding temperature fluctuation index threshold. If the temperature fluctuation index is greater than or equal to the temperature fluctuation index threshold, the temperature data in the distillation process causes the quality of the liquor to be unqualified, and the temperature data is adjusted; obtain the steam flow fluctuation index of the predicted unqualified liquor, and compare the steam flow fluctuation index with the corresponding steam flow fluctuation index threshold. If the steam flow fluctuation index is greater than or equal to the steam flow fluctuation index threshold, the steam flow data in the distillation process causes the quality of the liquor to be unqualified, and the steam flow data is adjusted; After adjusting the temperature data and steam flow data, the wine quality continues to be monitored and evaluated according to the time series.

Images