Baijiu solid state fermentation parameter online monitoring system and method

Abstract

The invention provides an online monitoring system and method for solid-state fermentation parameters of white spirit, and relates to the field of intelligent monitoring of white spirit fermentation. Through real-time analysis of temperature values and moisture content data at multiple time points in the Baijiu fermentation pit, whether fermentation abnormity exists in the pit or not is judged. Therefore, fermentation abnormity can be detected and judged in time, the intelligent level of the whole liquor fermentation process is improved, and the production efficiency and the product quality are improved.

Description

Technical Field

[0001] The present application relates to the field of intelligent monitoring of liquor fermentation, and more specifically, to an online monitoring system and method for liquor solid-state fermentation parameters. Background Art

[0002] As people's consumption level improves, consumers' demand for high-quality liquor is growing, prompting liquor manufacturers to continuously improve production processes and increase production efficiency through technological upgrades.

[0003] Solid-state fermentation is the core link of liquor production. In this link, it is particularly necessary to quickly and accurately obtain sensitive data such as cellar temperature and moisture for raw liquor production. Controlling the temperature and moisture inside the cellar is the key to producing high-quality liquor. The blending and filling links of liquor production enterprises have been automated, but the measurement of fermentation environment parameters is still at the physical and chemical measurement stage of manual sampling. This method easily destroys the integrity of the cellar fermentation environment, has low work efficiency, and cannot take into account the full cycle parameter monitoring of cellar fermentation.

[0004] Therefore, an optimized online monitoring scheme for liquor solid-state fermentation parameters is needed. Summary of the invention

[0005] In order to solve the above technical problems, the present application is proposed. The embodiments of the present application provide a system and method for online monitoring of liquor solid-state fermentation parameters.

[0006] According to one aspect of the present application, a system for online monitoring of liquor solid-state fermentation parameters is provided, comprising:

[0007] The data collection module for monitoring parameters of liquor solid-state fermentation is used to obtain the temperature and moisture content values ​​inside the cellar at multiple predetermined time points within a predetermined time period;

[0008] A liquor solid-state fermentation parameter monitoring data encoding module is used to perform feature encoding on the temperature values ​​and moisture content values ​​inside the cellar at the plurality of predetermined time points to obtain a cellar temperature-moisture correlation feature matrix;

[0009] The fermentation monitoring result generation module is used to determine whether there is fermentation abnormality in the cellar at the current time point based on the cellar temperature-moisture correlation characteristic matrix.

[0010] According to another aspect of the present application, a method for online monitoring of liquor solid-state fermentation parameters is provided, comprising:

[0011] Obtaining the temperature and moisture content values ​​inside the cellar at multiple predetermined time points within a predetermined time period;

[0012] Characteristic encoding is performed on the temperature values ​​and moisture content values ​​inside the cellar at the plurality of predetermined time points to obtain a cellar temperature-moisture correlation characteristic matrix;

[0013] Based on the cellar temperature-water correlation characteristic matrix, it is determined whether there is fermentation abnormality in the cellar at the current time point.

[0014] Compared with the prior art, the online monitoring system and method for solid-state fermentation parameters of liquor provided by the present application adopts data analysis technology based on deep learning, and performs real-time analysis of the temperature values ​​and moisture content data at multiple time points inside the liquor fermentation cellar to determine whether there is any fermentation abnormality in the cellar. In this way, fermentation abnormality can be detected and determined in a timely manner, which improves the intelligence level of the entire liquor fermentation process and is conducive to improving production efficiency and product quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0016] Figure 1 This is a system block diagram of the online monitoring system for solid-state fermentation parameters of liquor according to an embodiment of the present application.

[0017] Figure 2 It is a block diagram of the liquor solid-state fermentation parameter monitoring data encoding module in the liquor solid-state fermentation parameter online monitoring system according to the embodiment of the present application.

[0018] Figure 3 It is a block diagram of the liquor solid-state fermentation parameter monitoring data association unit in the liquor solid-state fermentation parameter online monitoring system according to an embodiment of the present application.

[0019] Figure 4 This is a block diagram of a fermentation monitoring result generation module in an online monitoring system for solid-state fermentation parameters of liquor according to an embodiment of the present application.

[0020] Figure 5 The present invention is a flow chart of a method for online monitoring of liquor solid-state fermentation parameters according to an embodiment of the present application. DETAILED DESCRIPTION

[0022] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0023] In today's society, with the increase in disposable income and the continuous improvement of the quality of life, consumers' demand for liquor tasting is no longer limited to simple drinking, but pays more attention to the quality and flavor of liquor. This change in consumption trend has directly promoted the liquor industry to move towards high-end and refined development. Liquor manufacturers have keenly captured this market change, and have increased their R&D investment, committed to improving the core competitiveness of their products through technological innovation.

[0024] Solid-state fermentation, as an indispensable soul process for brewing high-quality liquor, has become the focus of attention of enterprises for its meticulous management. In this complex and delicate fermentation process, every slight environmental change may have a profound impact on the flavor of the final liquor.

[0025] In actual liquor production, although liquor production companies have widely adopted automation and intelligent technologies in later processes such as blending and filling, achieving a leap in production efficiency, traditional manual sampling and physical and chemical testing methods still dominate the core link of solid-state fermentation. This method is not only cumbersome, time-consuming and labor-intensive, but more importantly, frequent sampling operations can easily break the delicate ecological balance inside the cellar, affecting the continuity and stability of the fermentation process. At the same time, manual monitoring is often difficult to achieve all-weather and full-cycle coverage, and it is easy to miss key information, resulting in a lack of timeliness and accuracy in the regulation of the fermentation process. Therefore, an online monitoring solution for liquor solid-state fermentation parameters is expected.

[0026] In recent years, deep learning and neural networks have been widely used in computer vision, natural language processing, text signal processing and other fields. In addition, deep learning and neural networks have also demonstrated a level close to or even beyond that of humans in image classification, object detection, semantic segmentation, text translation and other fields. The development of deep learning and neural networks has provided new solutions and solutions for the online monitoring of liquor solid-state fermentation parameters.

[0027] Figure 1 : is a system block diagram of the online monitoring system for liquor solid-state fermentation parameters according to an embodiment of the present application. Figure 1 As shown, in the online monitoring system 100 for solid-state fermentation parameters of liquor, it includes: a liquor solid-state fermentation parameter monitoring data collection module 110, which is used to obtain the temperature value and moisture content value inside the cellar at multiple predetermined time points within a predetermined time period; a liquor solid-state fermentation parameter monitoring data encoding module 120, which is used to feature encode the temperature value and moisture content value inside the cellar at the multiple predetermined time points to obtain a cellar temperature-moisture correlation feature matrix; a fermentation monitoring result generation module 130, which is used to determine whether there is fermentation abnormality in the cellar at the current time point based on the cellar temperature-moisture correlation feature matrix.

[0028] Specifically, in the technical scheme of the present application, first, the temperature value and moisture content value inside the cellar at multiple predetermined time points within a predetermined time period are obtained. It should be understood that temperature is a key factor affecting microbial activity and metabolic rate. Different microorganisms have different adaptability to temperature. Too high or too low temperature may inhibit the growth of microorganisms or cause their death, thereby affecting the normal progress of the fermentation process. Water is necessary for microbial growth and metabolism. An appropriate amount of water can ensure the normal activity of microorganisms and promote the progress of fermentation reactions. The content of water can also affect the solubility and diffusion rate of the fermentation substrate, thereby affecting the fermentation efficiency and product distribution. Therefore, in the solid-state fermentation process of liquor, temperature and moisture are two crucial environmental parameters, which directly affect the growth, metabolic activity and generation of fermentation products of microorganisms. By real-time monitoring of the temperature value and moisture content value inside the cellar, the state of the fermentation environment can be timely understood, and the basis for adjusting and controlling the fermentation process is provided for production personnel, thereby ensuring the stability of the fermentation process and the quality of liquor.

[0029] In an embodiment of the present application, a possible implementation method for obtaining the temperature value and moisture content value inside the cellar at multiple predetermined time points within a predetermined time period may be: the temperature data inside the cellar may be obtained by a temperature sensor installed in the cellar; the moisture data inside the cellar may be obtained by a moisture sensor installed in the cellar.

[0030] Figure 2 FIG. 1 is a block diagram of a liquor solid-state fermentation parameter monitoring data encoding module in a liquor solid-state fermentation parameter online monitoring system according to an embodiment of the present application. Figure 2 As shown, the liquor solid-state fermentation parameter monitoring data encoding module 120 includes: a liquor solid-state fermentation parameter monitoring data association unit 121, which is used to extract association features of the temperature values ​​and moisture content values ​​inside the cellar at the multiple predetermined time points to obtain a cellar temperature-moisture association matrix; and a liquor solid-state fermentation parameter monitoring data association feature generation unit 122, which is used to extract features of the cellar temperature-moisture association matrix to obtain the cellar temperature-moisture association feature matrix.

[0031] Figure 3 FIG. 1 is a block diagram of a liquor solid-state fermentation parameter monitoring data association unit in a liquor solid-state fermentation parameter online monitoring system according to an embodiment of the present application. Figure 3As shown, the liquor solid-state fermentation parameter monitoring data association unit 121 includes: a cellar temperature data structuring subunit 1211, which is used to arrange the temperature values ​​inside the cellar at the multiple predetermined time points according to the time dimension as a cellar temperature time series input vector; a cellar temperature time series feature generating subunit 1212, which is used to pass the cellar temperature time series input vector through a cellar monitoring parameter time series feature encoder to obtain a cellar temperature time series feature vector; a cellar moisture data structuring subunit 1213, which is used to arrange the moisture content values ​​inside the cellar at the multiple predetermined time points according to the time dimension as a cellar moisture content time series input vector; a cellar moisture time series feature generating subunit 1214, which is used to pass the cellar moisture content time series input vector through the cellar monitoring parameter time series feature encoder to obtain a cellar moisture content time series feature vector; a feature vector association subunit 1215, which is used to associate the cellar temperature time series feature vector with the cellar moisture content time series feature vector to obtain the cellar temperature-moisture association matrix.

[0032] Considering that the fermentation process is a typical time series process, the change of temperature parameters has obvious time series dependence, that is, the temperature value at the current moment is not only affected by the current environmental conditions and operations, but also closely related to the temperature value at the previous time point. Therefore, in the technical solution of the present application, the temperature values ​​inside the cellar at the multiple predetermined time points are arranged according to the time dimension as the cellar temperature time series input vector. By arranging according to the time dimension, this time series information can be completely retained, so that subsequent analysis and processing can make full use of these time series dependencies. Specifically, the change trend and fluctuation of temperature are often closely related to the change of fermentation state, such as the growth rate of microorganisms, the generation rate of metabolites, etc. Therefore, by analyzing the temperature time series data, it is possible to indirectly judge whether the fermentation state is normal.

[0033] In order to capture the key information in the temperature time series, such as the trend, periodicity, mutation point, etc. of temperature change, it is necessary to convert the original cellar temperature data into a higher-level and more meaningful representation, that is, to pass the cellar temperature time series input vector through the cellar monitoring parameter time series feature encoder to obtain the cellar temperature time series feature vector. The structure of the cellar monitoring parameter time series feature encoder in this application is a convolutional neural network model including a first convolution layer and a second convolution layer, wherein the first convolution layer uses a one-dimensional convolution kernel of a first scale, and the second convolution layer uses a one-dimensional convolution kernel of a second scale, and the first scale is different from the second scale. It should be understood that using one-dimensional convolution kernels of different scales can extract feature information at different time resolutions. The first convolution layer may focus more on local detail information, such as temperature fluctuations in a short period of time; while the second convolution layer may focus more on global trend information, such as the long-term trend of temperature changes. This multi-scale feature extraction helps to more comprehensively understand the inherent laws of the cellar temperature time series data, thereby improving the accuracy of abnormal cellar fermentation judgment.

[0034] In an embodiment of the present application, a possible implementation method of passing the cellar temperature time series input vector through a cellar monitoring parameter time series feature encoder to obtain a cellar temperature time series feature vector may be: inputting the cellar temperature time series input vector into the first convolution layer of the cellar monitoring parameter time series feature encoder to obtain a first-scale cellar temperature time series feature vector, wherein the first convolution layer has a one-dimensional convolution kernel of a first scale; inputting the cellar temperature time series input vector into the second convolution layer of the cellar monitoring parameter time series feature encoder to obtain a second-scale cellar temperature time series feature vector, wherein the second convolution layer has a one-dimensional convolution kernel of a second scale, and the first scale is different from the second scale; and cascading the first-scale cellar temperature time series feature vector and the second-scale cellar temperature time series feature vector to obtain the cellar temperature time series feature vector.

[0035] Moisture content is one of the important indicators that reflect the state of fermentation. During the fermentation process, if the moisture content is too high or too low, it may imply fermentation abnormalities. Too high a moisture content means that the fermentation liquid is too dilute, affecting the metabolic activity of the microorganisms; while too low a moisture content means that the fermentation liquid is too concentrated, affecting the generation and quality of the product. In order to clearly show the changing trend of moisture content over time and thus capture this dynamic change, in the technical solution of the present application, it is necessary to arrange the moisture content values ​​inside the cellar at the multiple predetermined time points according to the time dimension as a cellar moisture content time series input vector. By analyzing the cellar moisture content time series input vector, abnormal fluctuations in moisture content can be discovered in a timely manner. These abnormal fluctuations provide important data support for judging fermentation abnormalities.

[0036] The cellar moisture content time series input vector may contain noise caused by measurement errors, environmental factors or other non-fermentation process factors. In addition, there may be redundant information in the original data, which is of no substantial help for the subsequent fermentation situation judgment. In order to capture the key features of the moisture content change trend, it is necessary to pass the cellar moisture content time series input vector through the cellar monitoring parameter time series feature encoder to obtain the cellar moisture content time series feature vector. The cellar moisture content time series feature vector obtained by the encoder contains the moisture content change trend and the periodic characteristics of the moisture content.

[0037] The solid-state fermentation of liquor is a complex biochemical process, in which temperature and moisture are two crucial factors. Not only do they each have a significant impact on the fermentation process, but there are also complex interactions between them. Considering that analyzing one of the factors alone may not be able to fully reflect the fermentation state, the cellar temperature time series feature vector and the cellar moisture content time series feature vector are associated in the technical solution of the present application to obtain the cellar temperature-moisture association matrix. It should be understood that by associating the two feature vectors of temperature and moisture, the mutual influence between them can be comprehensively considered, so as to more accurately judge whether an abnormality occurs in the fermentation process. For example, in some cases, an abnormal increase in temperature may be accompanied by a rapid decrease in moisture, and this association information is crucial for judging the fermentation state. That is, through association analysis, the abnormal association pattern between temperature and moisture can be discovered in time, thereby warning of possible problems in the fermentation process in advance.

[0038] In an embodiment of the present application, a possible implementation method of associating the cellar temperature time series feature vector and the cellar moisture content time series feature vector to obtain the cellar temperature-moisture association matrix may be: calculating the cellar temperature-moisture association matrix between the cellar temperature time series feature vector and the cellar moisture content time series feature vector using the following association coding formula; wherein the association coding formula is:

[0039]

[0040] Where V a represents the cellar temperature time series feature vector, represents the transposed vector of the cellar temperature time series feature vector, V b represents the time series feature vector of the water content in the pit, M a represents the cellar temperature-moisture correlation matrix, Represents vector multiplication.

[0041] The cellar temperature-water association matrix obtained by association contains a large amount of redundant information and noise, which may contribute little to abnormal judgment or even interfere. It is necessary to perform feature extraction processing on it, which can effectively remove noise, thereby extracting the most critical association features for abnormal judgment, that is, passing the cellar temperature-water association matrix through a two-way attention mechanism module to obtain a cellar temperature-water association feature matrix. The process of feature extraction of the cellar temperature-water association matrix is ​​actually to abstract and generalize the original association matrix data and extract more representative features. These features may have better stability and consistency under different fermentation conditions, thereby enhancing the generalization ability of the model, so that it can more accurately judge fermentation anomalies under different circumstances. Those of ordinary skill in the art should know that the two-way attention mechanism module is an attention mechanism in deep learning, which can simultaneously consider context information when processing sequence data, thereby more accurately capturing key features in the data. The two-way attention mechanism module can more accurately capture abnormal patterns in the association matrix by accurately calculating the attention weight, thereby improving the accuracy of abnormal judgment. It is also worth mentioning that since the context information is considered at the same time, the two-way attention mechanism has stronger robustness to noise and outliers in the data.

[0042] In an embodiment of the present application, a possible implementation method of passing the cellar temperature-moisture association matrix through a bidirectional attention mechanism module to obtain a cellar temperature-moisture association feature matrix may be: pooling the cellar temperature-moisture association matrix along the horizontal direction and the vertical direction respectively to obtain a first-direction cellar temperature-moisture association pooling vector and a second-direction cellar temperature-moisture association pooling vector; performing association encoding on the first-direction cellar temperature-moisture association pooling vector and the second-direction cellar temperature-moisture association pooling vector to obtain a bidirectional cellar temperature-moisture association matrix; inputting the bidirectional cellar temperature-moisture association matrix into a Sigmoid activation function to obtain a bidirectional cellar temperature-moisture association weight matrix; calculating the position point multiplication between the bidirectional cellar temperature-moisture association weight matrix and the cellar temperature-moisture association matrix to obtain the cellar temperature-moisture association feature matrix.

[0043] Figure 4 FIG. 1 is a block diagram of a fermentation monitoring result generation module in an online monitoring system for solid-state fermentation parameters of liquor according to an embodiment of the present application. Figure 4As shown, the fermentation monitoring result generation module 130 includes: a feature matrix expansion unit 131, used to expand the cellar temperature-moisture correlation feature matrix to obtain a cellar temperature-moisture correlation feature vector; a feature optimization unit 132, used to suppress the interference distribution of the cellar temperature-moisture correlation feature vector based on intrinsic latent correlation analysis to obtain an optimized cellar temperature-moisture correlation feature vector; a feature vector classification unit 133, used to pass the optimized cellar temperature-moisture correlation feature vector through a fermentation identification classifier to obtain a classification result, and the classification result is used to indicate whether there is a fermentation abnormality in the cellar at the current time point.

[0044] Next, considering that the cellar temperature-water correlation feature matrix has a high dimension, which will increase the computational complexity and reduce the computational efficiency, in order to achieve dimensionality reduction of feature data, the cellar temperature-water correlation feature matrix is ​​expanded in the technical solution of the present application to obtain the cellar temperature-water correlation feature vector. By expanding the feature matrix, the two-dimensional data is converted into a one-dimensional vector, and in the process of expansion, the key information in the original feature matrix is ​​not lost, that is, it will not affect the judgment of whether there is fermentation abnormality in the cellar.

[0045] In particular, in the process of obtaining the internal temperature value and moisture content value of the cellar, the sensor may be affected by environmental noise, resulting in the collected data containing certain random errors, and the parameter selection and design of the cellar monitoring parameter time series feature encoder directly affect the quality of the feature vector. When the encoder is sensitive to noise and fails to fully capture the inherent laws of the data, unnecessary interference will be introduced into the cellar temperature-moisture correlation feature vector finally formed. Accordingly, in the technical solution of the present application, considering that the neural network model performs feature mining on the input cellar monitoring data based on the network parameter matrix, if the inherent implicit association between the network output feature and the network parameter matrix can be further considered, the prior information of the neural network model can be further used to modulate the characteristic manifold of the cellar temperature-moisture correlation feature vector to reduce the possibility of special instance disturbances of its eigenvalues. Therefore, the interference distribution suppression based on the inherent implicit association analysis is performed on the cellar temperature-moisture correlation feature vector to obtain the optimized cellar temperature-moisture correlation feature vector.

[0046] Specifically, in an embodiment of the present application, interference distribution suppression based on intrinsic implicit correlation analysis is performed on the cellar temperature-moisture correlation feature vector to obtain an optimized cellar temperature-moisture correlation feature vector, including: calculating the mean and variance of the cellar temperature-moisture correlation feature vector; calculating the difference between the eigenvalue of a predetermined position of the cellar temperature-moisture correlation feature vector and the mean of the cellar temperature-moisture correlation feature vector to obtain a first intermediate calculated value; multiplying the network parameter matrix by the cellar temperature-moisture correlation feature vector and then subtracting the cellar temperature-moisture correlation feature vector by position to obtain a first intermediate vector, and calculating the Frobenius norm of the first intermediate vector to obtain a first norm value; adding the first norm value to the cellar temperature-moisture correlation feature vector. The square root of the variance of the cellar temperature-moisture correlation characteristic vector is calculated to obtain a second intermediate calculation value; the first intermediate calculation value is divided by the second intermediate calculation value to obtain a third intermediate calculation value; the squares of each position of the cellar temperature-moisture correlation characteristic vector are calculated and then accumulated to obtain a first sum value; the square root of the first sum value is divided by the scale of the network parameter matrix and then the natural exponential function value is calculated to obtain a fourth intermediate calculation value, wherein the scale of the network parameter matrix is ​​equal to the product between the width value and the height value of the network parameter matrix; the fourth intermediate calculation value is multiplied by the weighted hyperparameter and then added to the third intermediate calculation value to obtain the characteristic value of the predetermined position of the optimized cellar temperature-moisture correlation characteristic vector.

[0047] In an embodiment of the present application, another expression of performing interference distribution suppression based on intrinsic latent correlation analysis on the cellar temperature-moisture correlation feature vector to obtain the optimized cellar temperature-moisture correlation feature vector may be: processing the cellar temperature-moisture correlation feature vector with the following formula to obtain the optimized cellar temperature-moisture correlation feature vector; wherein the formula is:

[0048]

[0049] Where V is the cellar temperature-water association feature vector, M is the network parameter matrix, β represents the weighted hyperparameter, and v i is the eigenvalue of the ith position of the cellar temperature-water correlation eigenvector, N is the scale of the network parameter matrix, the scale of the network parameter matrix is ​​equal to the product of the width value and the height value of the network parameter matrix, L is the length of the cellar temperature-water correlation eigenvector, ||·|| F represents the F norm of the eigenvector, μ represents the mean of the cellar temperature-water correlation eigenvector, σ 2 represents the variance of the cellar temperature-water correlation characteristic vector, exp(·) represents the exponential function with the natural constant e as the base, v'i represents the eigenvalue of the i-th position of the optimized cellar temperature-water correlation eigenvector, Represents matrix multiplication.

[0050] It should be understood that in the technical solution of the present application, considering that when the neural network model performs feature mining on the input data based on the network parameter matrix, if the inherent implicit association between the network output feature and the network parameter matrix can be further considered, the prior information of the neural network model can be further used to modulate the characteristic manifold of the cellar temperature-water association feature vector to reduce the possibility of the occurrence of special instance disturbances of its characteristic values. Based on this, in a preferred embodiment, the cellar temperature-water association feature vector is subjected to interference distribution suppression based on the inherent implicit association analysis, which first maps the cellar temperature-water association feature vector to the high-dimensional feature space where the network parameter matrix is ​​located, and uses the compatibility perception factor between the mapping result and the original input as the inherent implicit association descriptor between the network output feature and the network parameter matrix. Then, the statistical characteristics of the feature group of the cellar temperature-water association feature vector are used as the inherent modal information of the cellar temperature-water association feature vector, and the inherent implicit association descriptor is used as the instance disturbance differential adjustment variable, and the attribute values ​​of each position of the cellar temperature-water association feature vector are subjected to data dynamic standardization. At the same time, the class-standardized modulus data of the cellar temperature-moisture correlation feature vector is used as a bias item to constrain the acceptable range of the population distribution of the cellar temperature-moisture correlation feature vector. In this way, the attribute influence of the interference distribution of abnormal samples can be suppressed as much as possible, thereby improving the expression accuracy of the cellar temperature-moisture correlation feature vector.

[0051] Finally, the optimized cellar temperature-water association feature vector is passed through the fermentation identification classifier to obtain a classification result, and the classification result is used to indicate whether there is a fermentation abnormality in the cellar at the current time point. It should be understood that although the optimized cellar temperature-water association feature vector contains the monitoring situation in the cellar, it is information in a complex high-dimensional space and does not directly represent the situation in the fermentation tank. The fermentation identification classifier is a machine learning model for distinguishing between normal and abnormal states in the fermentation process. It learns the characteristic patterns under the normal fermentation state by training a large amount of historical fermentation data, and can classify and judge the fermentation state according to these characteristic patterns when new monitoring data arrives. When the classification result shows that there is a fermentation abnormality in the cellar at the current time point, the early warning mechanism can be triggered immediately to notify relevant personnel to handle it. This helps to discover and solve problems in the fermentation process in a timely manner and avoid further expansion of abnormal situations. By introducing the fermentation identification classifier, the automatic detection and judgment of fermentation abnormalities can be realized, reducing the influence of manual intervention and subjective judgment. This helps to improve the intelligence level of the entire fermentation process and improve production efficiency and product quality.

[0052] In an embodiment of the present application, the optimized cellar temperature-water association feature vector is passed through a fermentation identification classifier to obtain a classification result, and the classification result is used to indicate whether there is a fermentation abnormality in the cellar at the current time point. One possible implementation method may be: using the fermentation identification classifier to process the optimized cellar temperature-water association feature vector using the following classification formula to generate the classification result; wherein the classification formula is:

[0053] O=softmax{(M c ,B c )|V c}

[0054] Wherein O is the classification result, V c represents the optimized cellar temperature-water correlation feature vector, M c is the weight matrix of the fully connected layer, B c Represents the bias vector of the fully connected layer, and softmax is the normalized exponential function.

[0055] In summary, the online monitoring system 100 for solid-state fermentation parameters of liquor based on the embodiment of the present application is explained, which uses data analysis technology based on deep learning to perform real-time analysis of temperature values ​​and moisture content data at multiple time points inside the liquor fermentation cellar to determine whether there is fermentation abnormality in the cellar. In this way, fermentation abnormality can be detected and determined in a timely manner, which improves the intelligence level of the entire liquor fermentation process and is conducive to improving production efficiency and product quality.

[0056] As described above, the online monitoring system 100 for solid-state fermentation parameters of liquor according to the embodiment of the present application can be implemented in various terminal devices, such as a server for online monitoring of solid-state fermentation parameters of liquor. In one example, the online monitoring system 100 for solid-state fermentation parameters of liquor according to the embodiment of the present application can be integrated into the terminal device as a software module and / or a hardware module. For example, the online monitoring system 100 for solid-state fermentation parameters of liquor can be a software module in the operating system of the terminal device, or can be an application developed for the terminal device; of course, the online monitoring system 100 for solid-state fermentation parameters of liquor can also be one of the many hardware modules of the terminal device.

[0057] Alternatively, in another example, the white wine solid-state fermentation parameter online monitoring system 100 and the terminal device may also be separate devices, and the white wine solid-state fermentation parameter online monitoring system 100 may be connected to the terminal device via a wired and / or wireless network, and transmit interactive information in accordance with an agreed data format.

[0058] Figure 5 Flow chart of the method for online monitoring of liquor solid-state fermentation parameters according to an embodiment of the present application. Figure 5 As shown, in the online monitoring method of liquor solid-state fermentation parameters, the method includes: S110, obtaining the temperature value and moisture content value inside the cellar at multiple predetermined time points within a predetermined time period; S120, feature encoding the temperature value and moisture content value inside the cellar at the multiple predetermined time points to obtain a cellar temperature-moisture correlation feature matrix; S130, based on the cellar temperature-moisture correlation feature matrix, judging whether there is fermentation abnormality in the cellar at the current time point.

[0059] Here, those skilled in the art will appreciate that the specific operations of each step in the above-mentioned method for online monitoring of liquor solid-state fermentation parameters have been described in the above reference. Figures 1 to 4 The online monitoring system for solid-state fermentation parameters of liquor has been introduced in detail in the description of the online monitoring system for solid-state fermentation parameters of liquor, and therefore, its repeated description will be omitted.

[0060] In summary, the online monitoring method of liquor solid-state fermentation parameters based on the embodiment of the present application is explained, which uses data analysis technology based on deep learning to perform real-time analysis of temperature values ​​and moisture content data at multiple time points inside the liquor fermentation cellar to determine whether there is fermentation abnormality in the cellar. In this way, fermentation abnormality can be detected and judged in a timely manner, which improves the intelligence level of the entire liquor fermentation process and is conducive to improving production efficiency and product quality.

[0061] Finally, the present application also relates to a computer-readable storage medium for realizing the online monitoring method of liquor solid-state fermentation parameters according to one or more embodiments of the present application. The computer-readable storage medium mentioned herein includes various types of computer storage media, which can be any available medium that can be accessed by a general or special-purpose computer. For example, the computer-readable storage medium may include RAM, ROM, EPROM, E2PROM, register, hard disk, removable disk, CD-ROM or other optical disk storage, disk storage or other magnetic storage device, or any other temporary or non-temporary medium that can be used to carry or store a desired program code unit in the form of an instruction or data structure and can be accessed by a general or special-purpose computer, or a general or special-purpose processor. The description of the computer-readable storage medium according to the present application can refer to the explanation of the online monitoring method of liquor solid-state fermentation parameters according to the present application, which will not be repeated here.

[0062] It is obvious to those skilled in the art that the present application is not limited to the details of the above exemplary embodiments, and that the present application can be implemented in other specific forms without departing from the spirit or essential features of the present application.

[0063] Therefore, no matter from which point of view, the embodiments should be regarded as illustrative and non-restrictive, and the scope of the present application is limited by the appended claims rather than the above description, so it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims are included in the present application. Any attached figure mark in the claims should not be regarded as limiting the claims involved.

[0064] In addition, it is clear that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units or devices stated in the system claim can also be implemented by one unit or device through software or hardware. The second and other words are used to indicate names, but not to indicate any particular order.

[0065] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present application and are not intended to limit it. Although the present application has been described in detail with reference to the preferred embodiments, a person of ordinary skill in the art should understand that the technical solution of the present application may be modified or replaced by equivalents without departing from the spirit of the technical solution of the present application.

Claims

1. An online monitoring system for liquor solid-state fermentation parameters, characterized in that: include: The data collection module for monitoring parameters of liquor solid-state fermentation is used to obtain the temperature and moisture content values ​​inside the cellar at multiple predetermined time points within a predetermined time period; A liquor solid-state fermentation parameter monitoring data encoding module is used to perform feature encoding on the temperature values ​​and moisture content values ​​inside the cellar at the plurality of predetermined time points to obtain a cellar temperature-moisture correlation feature matrix; The fermentation monitoring result generation module is used to determine whether there is fermentation abnormality in the cellar at the current time point based on the cellar temperature-moisture correlation characteristic matrix.

2. The on-line monitoring system for liquor solid-state fermentation parameters according to claim 1, characterized in that: The liquor solid-state fermentation parameter monitoring data encoding module comprises: A liquor solid-state fermentation parameter monitoring data association unit is used to extract association features of the temperature values ​​and moisture content values ​​inside the cellar at the plurality of predetermined time points to obtain a cellar temperature-moisture association matrix; The liquor solid-state fermentation parameter monitoring data association feature generation unit is used to extract features from the cellar temperature-moisture association matrix to obtain the cellar temperature-moisture association feature matrix.

3. The online monitoring system for liquor solid-state fermentation parameters according to claim 2 is characterized in that: The liquor solid-state fermentation parameter monitoring data association unit comprises: The cellar temperature data structuring subunit is used to arrange the temperature values ​​inside the cellar at the plurality of predetermined time points into a cellar temperature time series input vector according to the time dimension; A cellar temperature time series feature generating subunit, used for passing the cellar temperature time series input vector through a cellar monitoring parameter time series feature encoder to obtain a cellar temperature time series feature vector; The cellar moisture data structuring subunit is used to arrange the moisture content values ​​inside the cellar at the plurality of predetermined time points according to the time dimension into a cellar moisture content time series input vector; A cellar water content time series feature generating subunit, used for passing the cellar water content time series input vector through the cellar monitoring parameter time series feature encoder to obtain a cellar water content time series feature vector; The feature vector association subunit is used to associate the cellar temperature time series feature vector with the cellar moisture content time series feature vector to obtain the cellar temperature-moisture association matrix.

4. The on-line monitoring system for liquor solid-state fermentation parameters according to claim 3 is characterized in that: The liquor solid-state fermentation parameter monitoring data association feature generation unit is used to: obtain the cellar temperature-moisture association feature matrix by passing the cellar temperature-moisture association matrix through a two-way attention mechanism module.

5. The on-line monitoring system for liquor solid-state fermentation parameters according to claim 4 is characterized in that: The cellar monitoring parameter temporal feature encoder is a convolutional neural network model comprising a first convolutional layer and a second convolutional layer, wherein the first convolutional layer uses a one-dimensional convolutional kernel of a first scale, and the second convolutional layer uses a one-dimensional convolutional kernel of a second scale, and the first scale is different from the second scale.

6. The on-line monitoring system for liquor solid-state fermentation parameters according to claim 5 is characterized in that: The fermentation monitoring result generating module comprises: A characteristic matrix expansion unit, used for expanding the cellar temperature-water correlation characteristic matrix to obtain a cellar temperature-water correlation characteristic vector; A feature optimization unit, used for performing interference distribution suppression on the cellar temperature-moisture correlation feature vector based on intrinsic latent correlation analysis to obtain an optimized cellar temperature-moisture correlation feature vector; The feature vector classification unit is used to pass the optimized cellar temperature-water association feature vector through a fermentation identification classifier to obtain a classification result, and the classification result is used to indicate whether there is a fermentation abnormality in the cellar at the current time point.

7. The on-line monitoring system for liquor solid-state fermentation parameters according to claim 6, characterized in that: The feature optimization unit comprises: Calculating the mean and variance of the cellar temperature-water correlation feature vector; Calculating the difference between the characteristic value of the predetermined position of the cellar temperature-moisture correlation characteristic vector and the mean value of the cellar temperature-moisture correlation characteristic vector to obtain a first intermediate calculation value; Multiplying the network parameter matrix by the cellar temperature-moisture correlation feature vector and then subtracting the network parameter matrix from the cellar temperature-moisture correlation feature vector by position to obtain a first intermediate vector, and calculating the Frobenius norm of the first intermediate vector to obtain a first norm value; The first norm value is added to the variance of the cellar temperature-water correlation characteristic vector and then a square root operation is performed to obtain a second intermediate calculated value; dividing the first intermediate calculated value by the second intermediate calculated value to obtain a third intermediate calculated value; Calculating the square of each position of the cellar temperature-water correlation characteristic vector and then accumulating and summing them to obtain a first sum value; Calculate a natural exponential function value after taking the square root of the first sum value and dividing it by the scale of the network parameter matrix to obtain a fourth intermediate calculation value, wherein the scale of the network parameter matrix is ​​equal to the product of the width value of the network parameter matrix and the height value thereof; The fourth intermediate calculated value is multiplied by the weighted hyperparameter and then added to the third intermediate calculated value to obtain the characteristic value of the predetermined position of the optimized cellar temperature-water correlation characteristic vector.

8. A method for online monitoring of liquor solid-state fermentation parameters, characterized in that: include: Obtaining the temperature and moisture content values ​​inside the cellar at multiple predetermined time points within a predetermined time period; Characteristic encoding is performed on the temperature values ​​and moisture content values ​​inside the cellar at the plurality of predetermined time points to obtain a cellar temperature-moisture correlation characteristic matrix; Based on the cellar temperature-water correlation characteristic matrix, it is determined whether there is fermentation abnormality in the cellar at the current time point.

9. The method for online monitoring of liquor solid-state fermentation parameters according to claim 8, characterized in that: The temperature values ​​and moisture content values ​​inside the cellar at the plurality of predetermined time points are feature-coded to obtain a cellar temperature-moisture correlation feature matrix, including: Extracting correlation features of the temperature values ​​and moisture content values ​​inside the cellar at the plurality of predetermined time points to obtain a cellar temperature-moisture correlation matrix; Feature extraction is performed on the cellar temperature-moisture correlation matrix to obtain the cellar temperature-moisture correlation feature matrix.

10. The method for online monitoring of liquor solid-state fermentation parameters according to claim 9, characterized in that: Performing feature extraction on the cellar temperature-water correlation matrix to obtain the cellar temperature-water correlation feature matrix includes: Arranging the temperature values ​​inside the cellar at the plurality of predetermined time points into a cellar temperature time series input vector according to the time dimension; Passing the cellar temperature time series input vector through a cellar monitoring parameter time series feature encoder to obtain a cellar temperature time series feature vector; Arranging the moisture content values ​​inside the cellar at the plurality of predetermined time points according to the time dimension into a cellar moisture content time series input vector; Passing the cellar water content time series input vector through the cellar monitoring parameter time series feature encoder to obtain a cellar water content time series feature vector; The cellar temperature time series feature vector and the cellar moisture content time series feature vector are associated to obtain the cellar temperature-moisture association matrix.

Images