Intelligent fermentation tank group management method and system based on optimal fermentation time of soy sauce

Abstract

Disclosed in the present invention are an intelligent fermentation tank group management method and system based on the optimal fermentation time of soy sauce. The method comprises: collecting production data of a fermentation tank group and storing same as historical production data of the fermentation tank group; preprocessing the historical production data of the fermentation tank group; on the basis of the preprocessed historical production data, calculating a correlation coefficient between fermentation process data and product quality indexes; on the basis of the preprocessed historical production data, establishing a neural network model for a mathematical relationship between the fermentation process data and product quality index data; using the preprocessed historical production data as sample data, and using a gradient descent algorithm to train the neural network model; on the basis of the fermentation process data of the fermentation tank group collected in real time, using the trained neural network model to output predicted product quality index data of the fermentation tank group; and on the basis of the product quality index data, establishing a dynamic management mathematical model, with the objective of optimizing the optimal fermentation time of soy sauce and the overall production quality of the fermentation tank group, and using a genetic algorithm to calculate a solution of the optimization objective, so as to realize intelligent management of the fermentation tank group. In the present invention, big data, a neural network model, and a genetic algorithm are combined, and the overall optimization of quality indexes of products from fermentation tank groups is achieved by calculating the optimal fermentation time of soy sauce, so as to realize intelligent management of fermentation tank groups.

Description

Intelligent Management Method and System for Fermentation Tank Groups Based on Optimal Fermentation Time for Soy Sauce Technical Field

[0001] This invention belongs to the field of intelligent management technology for food processing, and in particular relates to an intelligent management method, system, terminal equipment, and computer-readable storage medium for a group of fermentation tanks based on the optimal fermentation time for soy sauce. Background Technology

[0002] Modern soy sauce brewing technology is based on traditional soy sauce fermentation processes, and achieves industrialized production of soy sauce through mechanization and automation of the production process.

[0003] Fermentation is a crucial and lengthy process in soy sauce production. Due to the long fermentation time, soy sauce manufacturers need to build a large number of fermentation tanks, and in daily production, they need to dynamically manage all of these tanks.

[0004] Currently, most soy sauce producers use a fixed fermentation cycle and regular testing of product quality indicators to determine the total fermentation time for each fermentation tank and manage the fermentation tank group. However, different seasons and years, as well as factors such as temperature, light, and cycle time, can affect the fermentation effect of each tank, resulting in variations in the fermentation process. Using a fixed fermentation cycle will lead to differences in the quality indicators of the fermented products. While regular testing can ensure that product quality indicators exceed production requirements, it is still difficult to achieve overall optimization of the quality indicators of products produced by the entire fermentation tank group.

[0005] Summary of the Invention

[0006] To address the shortcomings of the existing technology, this invention provides an intelligent management method, system, terminal equipment, and computer-readable storage medium for fermentation tank groups based on the optimal fermentation time of soy sauce. It combines big data, neural network models, and genetic algorithms to optimize the overall quality indicators of products produced by the fermentation tank group by calculating the optimal fermentation time of soy sauce, thereby achieving intelligent management of the fermentation tank group.

[0007] The first objective of this invention is to provide an intelligent management method for a group of fermentation tanks based on the optimal fermentation time for soy sauce.

[0008] The second objective of this invention is to provide an intelligent management system for a group of fermentation tanks based on the optimal fermentation time for soy sauce.

[0009] The third objective of this invention is to provide a terminal device.

[0010] A fourth objective of this invention is to provide a computer-readable storage medium.

[0011] The first objective of this invention can be achieved by adopting the following technical solution:

[0012] A method for intelligent management of fermentation tank groups based on the optimal fermentation time for soy sauce, the method comprising:

[0013] Collect and store the production data of the fermentation tank group as historical production data of the fermentation tank group; the production data includes fermentation process data and product quality index data;

[0014] Preprocess the historical production data of the fermentation tank group;

[0015] Based on the pretreated historical production data, the correlation coefficient between fermentation process data and product quality indicators was calculated.

[0016] Based on preprocessed historical production data, a neural network model is established to establish the mathematical relationship between fermentation process data and product quality index data. The preprocessed historical production data is used as sample data, and the neural network model is trained using the gradient descent algorithm. When selecting the value of the gradient descent algorithm, the iteration calculation step size is selected according to the correlation coefficient between the fermentation process data and the product quality index data to improve the calculation efficiency.

[0017] Based on the real-time collected fermentation process data of the fermentation tank group, the trained neural network model is used to output the predicted product quality index data of the fermentation tank group. Based on the product quality index data, with the goal of optimizing the optimal fermentation time of soy sauce and the overall production quality of the fermentation tank group, a dynamic management mathematical model is established, and a genetic algorithm is used to calculate the solution of the optimization goal, so as to realize the intelligent management of the fermentation tank group.

[0018] Furthermore, the product quality index data includes the ammonia nitrogen content (ANN) and total nitrogen content (TNN) of soy sauce; the dynamic management mathematical model is:

[0019] In the formula, g1 and g2 are the ANN quality index and TNN quality index of the fermenters participating in the genetic algorithm calculation, respectively, n is the total number of fermenters participating in the genetic algorithm calculation, and ANN... i TNN i , respectively, represent the predicted ANN and TNN of the i-th fermenter; g3 is the overall product quality index of the fermenter group participating in the genetic algorithm calculation; a and b are the weight coefficients of the total predicted ANN and TNN, respectively.

[0020] Furthermore, in the genetic algorithm, the fermenters participating in the genetic algorithm calculation are called chromosomes, and the total number of all fermenters participating in the genetic algorithm calculation is the number of chromosomes.

[0021] Furthermore, the step of using preprocessed historical production data as sample data and training a neural network model using the gradient descent algorithm includes:

[0022] The fermentation cycle of soy sauce is divided into three independent fermentation stages: the initial fermentation stage, the middle fermentation stage, and the final fermentation stage. Each fermentation stage corresponds to a neural network model, that is, a neural network model is trained using historical production data of a fermentation stage.

[0023] The three neural network models are constructed and trained using the same method. We will use one of the neural network models as an example:

[0024] The input layer neurons are fermentation process data from preprocessed historical production data;

[0025] The output layer neurons contain predicted product quality index data.

[0026] The squared difference between the predicted product quality index data and the product quality index data in the preprocessed historical production data is used as the loss function, and the gradient descent algorithm is used to train the neural network model.

[0027] Furthermore, the step of using a trained neural network model to output predicted product quality index data of the fermentation tank group based on real-time collected fermentation process data of the fermentation tank group includes:

[0028] Based on the real-time data collection time, the fermentation stage is determined, and then the corresponding trained neural network model is determined.

[0029] After preprocessing the fermentation process data of the real-time collected fermentation tank group, the data is input into a defined neural network model, and the predicted product quality index data is output.

[0030] Furthermore, the fermentation process data includes the temperature inside the fermentation tank, light intensity, atmospheric temperature, atmospheric humidity, production date, and the row and column position of the fermentation tank in the tank area; the product quality index data includes the ammonia nitrogen content and total nitrogen content of the soy sauce.

[0031] The preprocessing includes data cleaning and data standardization, wherein:

[0032] Data cleaning includes: removing data where ammonia nitrogen is negatively correlated with time, as well as data points related to temperature, humidity, and light conditions that do not conform to natural laws;

[0033] Data standardization processing includes: integrating the daily fermentation tank temperature, light intensity, atmospheric temperature, and atmospheric humidity for each fermenter; and classifying and quantifying the row and column positions for each fermenter.

[0034] Furthermore, the step of selecting the iterative calculation step size based on the correlation coefficient between fermentation process data and product quality index data includes:

[0035] When determining the values ​​for the gradient descent algorithm, an appropriate step size should be selected for the step values ​​of the preprocessed fermentation process data: if the absolute value of the correlation coefficient is small, the step size should be increased appropriately; if the absolute value of the correlation coefficient is large, the step size should be decreased appropriately.

[0036] Furthermore, the correlation coefficient is calculated using the Spearman rank correlation coefficient.

[0037] The second objective of this invention can be achieved by adopting the following technical solution:

[0038] A smart management system for a group of fermentation tanks based on the optimal fermentation time for soy sauce, the system comprising:

[0039] The data acquisition module is used to collect and store the production data of the fermentation tank group as historical production data of the fermentation tank group; the production data includes fermentation process data and product quality index data.

[0040] The data preprocessing module is used to preprocess the historical production data of the fermentation tank group;

[0041] The correlation coefficient calculation module is used to calculate the correlation coefficient between fermentation process data and product quality indicators based on pre-processed historical production data.

[0042] The model training module is used to establish a neural network model of the mathematical relationship between fermentation process data and product quality index data based on preprocessed historical production data. The preprocessed historical production data is used as sample data, and the gradient descent algorithm is used to train the neural network model. When selecting the value of the gradient descent algorithm, the iteration calculation step size is selected according to the correlation coefficient between fermentation process data and product quality index data to improve computational efficiency.

[0043] The dynamic management module is used to output predicted product quality index data of the fermentation tank group based on the real-time collected fermentation process data of the fermentation tank group and a trained neural network model. Based on the product quality index data, with the goal of optimizing the optimal fermentation time of soy sauce and the overall production quality of the fermentation tank group, a dynamic management mathematical model is established, and the solution of the optimization objective is calculated by using a genetic algorithm to realize the intelligent management of the fermentation tank group.

[0044] The third objective of this invention can be achieved by adopting the following technical solution:

[0045] A terminal device includes a processor and a memory for storing processor-executable programs. When the processor executes the program stored in the memory, it implements the above-mentioned intelligent management method for fermentation tank groups based on the optimal fermentation time of soy sauce.

[0046] The fourth objective of this invention can be achieved by adopting the following technical solution:

[0047] A computer-readable storage medium storing a program that, when executed by a processor, implements the above-described intelligent management method for a group of fermentation tanks based on the optimal fermentation time for soy sauce.

[0048] The present invention has the following advantages over the prior art:

[0049] This invention utilizes an intelligent algorithm that combines big data, neural network models, and genetic algorithms to calculate the optimal fermentation time for soy sauce, thereby achieving the best overall product quality indicators for the fermentation tank group and maximizing the production efficiency of the fermentation tank group. Attached Figure Description

[0050] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0051] Figure 1 is a flowchart of the intelligent management method for fermentation tank groups based on the optimal fermentation time of soy sauce according to Embodiment 1 of the present invention;

[0052] Figure 2 is a schematic diagram of the intelligent management method for fermentation tank groups based on the optimal fermentation time of soy sauce according to Embodiment 1 of the present invention;

[0053] Figure 3 is a schematic diagram of the neural network model of Embodiment 1 of the present invention;

[0054] Figure 4 is a flowchart of the genetic algorithm for dynamic management of fermenter group in Embodiment 1 of the present invention;

[0055] Figure 5 is a structural block diagram of the intelligent management system for fermentation tank group based on the optimal fermentation time of soy sauce according to Embodiment 2 of the present invention;

[0056] Figure 6 is a structural block diagram of the terminal device of Embodiment 3 of the present invention. Detailed Implementation

[0057] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. It should be understood that the specific embodiments described are merely used to explain this application and are not intended to limit this application.

[0058] Example 1:

[0059] This embodiment provides an intelligent management method for fermentation tank groups based on the optimal fermentation time for soy sauce. First, historical production data of the fermentation tank group is acquired and preprocessed, including cleaning, conversion, and standardization. Then, using Spearman's rank correlation coefficient, the correlation between the quality indicators of the fermented product and the fermentation process data is calculated to analyze the impact of each process data on the quality indicators of the fermented product. Next, a neural network is used to calculate the mathematical relationship between each process data and the product quality indicators. With the goal of optimizing the optimal fermentation time for soy sauce and the overall production quality of the fermentation tank group, a mathematical model is established. Combining the real-time collected production data, process data, and the mathematical relationship between the product quality indicators of the fermentation tank group, a genetic algorithm is used to calculate the solution to the optimization objective, thereby achieving intelligent management of the fermentation tank group.

[0060] As shown in Figures 1 and 2, the intelligent management method for fermentation tank groups based on the optimal fermentation time of soy sauce provided in this embodiment includes the following steps:

[0061] S101. Collect and store the production data of the fermentation tank group as historical production data of the fermentation tank group; the production data includes fermentation process data and product quality index data.

[0062] The data that needs to be collected during the fermentation tank group production process includes:

[0063] 1)T I —Temperature inside the fermenter, detecting the material temperature at different times during the fermentation process in the fermenter;

[0064] 2) I—Light intensity, measuring the intensity of sunlight at different times during the fermentation process in the fermenter;

[0065] 3)T o —Atmospheric temperature, detecting the atmospheric temperature at different times during the fermentation process in the fermenter;

[0066] 4) RH—Atmospheric humidity, measuring the atmospheric humidity at different times during the fermentation process in the fermenter;

[0067] 5) D—Production Date, recording the date of different stages of the fermentation process in the fermenter;

[0068] 6) ANN—ammonia nitrogen content in soy sauce, one of the important indicators of product quality of fermented soy sauce;

[0069] 7) TNN—Total nitrogen content in soy sauce, one of the important indicators of product quality of fermented soy sauce.

[0070] The computer-controlled data acquisition system collects real-time data on the production process of the fermenters and stores it in a database, which serves as the basis for big data analysis of the fermenter group production.

[0071] In addition to collecting the production process data of the aforementioned fermentation tank group, the data acquisition system also needs to record the following data:

[0072] 1) PR—The row position of the fermenter within the tank area;

[0073] 2) PC—Column position of the tank area where the fermenter is located.

[0074] S102. Preprocess the historical production data of the fermentation tank group.

[0075] Before conducting data analysis, the historical production data of the fermentation tank group needs to be preprocessed to meet the requirements of the analysis and calculation. Data preprocessing includes:

[0076] 1) Data cleaning.

[0077] Remove data points where ammonia nitrogen is negatively correlated with time, and data points related to temperature, humidity, and light conditions that do not conform to natural laws.

[0078] The cleaned ANN and TNN are preprocessed product quality index data.

[0079] 2) Data standardization processing.

[0080] For each fermenter, T per day I The (temperature inside the fermenter) data is integrated to obtain ∫T I ;

[0081] Integrate the daily I (light intensity) data for each fermenter to obtain ∫I;

[0082] For each fermenter, T per day o Integrating the (atmospheric temperature) data yields ∫T o ;

[0083] Integrate the daily RH (atmospheric humidity) data for each fermenter to obtain ∫RH;

[0084] The PR (row position) and PC (column position) data of each fermenter are classified and quantified to obtain the PR-PC parameters of the fermenter.

[0085] ∫T I 、∫I、∫T o ∫RH, D, and PR-PC represent the fermentation process data after pretreatment.

[0086] 3) Utilize big data to establish an annual database of temperature, humidity, and light intensity as the basis for trend data prediction.

[0087] S103. Based on the historical production data after pretreatment, calculate the correlation coefficient between the fermentation process data and the product quality indicators.

[0088] The impact of each fermentation process data on product quality indicators varies. By calculating the Spearman rank correlation coefficient, the magnitude of the impact of each fermentation process data on product quality indicators can be analyzed.

[0089] ∫T I 、∫I、∫T o ∫RH, D, and PR-PC are paired with ANN and TNN respectively, and the correlation coefficient ∫T is calculated using the Spearman rank correlation coefficient formula. I 、∫I、∫T o Calculated values ​​of the Spearman rank correlation coefficients between the parameters ∫RH, D, and PR-PC and ANN and TNN.

[0090] The formula for calculating the Spearman rank correlation coefficient is as follows:

[0091] In the formula:

[0092] d i —The difference in rank between the i-th pairs of data;

[0093] n — the total number of pairs.

[0094] The correlation analysis between data from each fermentation process and product quality indicators is based on the following:

[0095] Positive value: When the Spearman rank correlation coefficient is positive, it indicates that there is a positive correlation between the ranks of the two variables, that is, if the rank of one variable increases, the rank of the other variable also tends to increase.

[0096] Negative value: When the Spearman rank correlation coefficient is negative, it indicates that there is a negative correlation between the ranks of the two variables, that is, if the rank of one variable increases, the rank of the other variable tends to decrease.

[0097] Zero value: When the Spearman rank correlation coefficient is zero, it means that there is no monotonic relationship between the ranks of the two variables.

[0098] The Spearman rank correlation coefficient calculation provides an important reference for fitting the mathematical relationship between subsequent fermentation production process data and product quality indicators, reducing the complexity of fitting calculations and improving the system's computational efficiency.

[0099] S104. Using the preprocessed historical production data as sample data, train the neural network model using the gradient descent algorithm; when selecting values ​​for the gradient descent algorithm, select the iteration calculation step size based on the correlation coefficient.

[0100] The pre-treated fermentation process data of each fermenter is input into the neural network model every day, and the neural network model outputs predicted product quality index data.

[0101] Based on the historical production data after pretreatment, the fermentation process data (∫T) were determined. I 、∫I、∫T o The mathematical relationship between ∫RH, D, PR-PC and product quality index data (ANN, TNN) is a crucial and computationally intensive aspect of intelligent management of fermentation tank groups. The trained neural network model can determine the fermentation process data (∫T... I 、∫I、∫T o Mathematical relationships between ∫RH, D, PR-PC and product quality index data (ANN, TNN).

[0102] The mathematical relationship between fermentation process data and product quality data varies significantly at different stages of soy sauce fermentation. If a single formula were used to represent the mathematical relationship between fermentation process data and product quality data for the entire fermentation cycle of soy sauce in a fermentation tank, the formula would be extremely complex and its effectiveness would not be guaranteed.

[0103] In this embodiment, the soy sauce fermentation cycle is divided into three independent stages: the initial fermentation stage, the middle fermentation stage, and the final fermentation stage, which account for approximately 1 / 4, 1 / 2, and 1 / 4 of the complete fermentation cycle, respectively, calculated in days. Three neural network models (with different mathematical formulas) are used to represent the mathematical relationships between the fermentation process data and product quality data for the three stages.

[0104] The three neural network models are constructed and trained using the same method. The following explanation uses one of them as an example:

[0105] 1) Establish a neural network model to establish the relationship between fermentation process data and product quality index data.

[0106] Establish the mathematical expression for the neuron:

[0107] In the formula:

[0108] —The j-th neuron in the l-th layer;

[0109] n (l-1) —The number of neurons in the (l-1)th layer;

[0110] —The weights from the i-th neuron in layer (l-1) to the j-th neuron in layer l;

[0111] —The output of the i-th neuron in the (l-1)-th layer;

[0112] —The bias term of the j-th neuron in the l-th layer;

[0113] g() — Activation function.

[0114] As shown in Figure 3, a two-layer neural network is used to represent the mathematical relationship between fermentation process data and product quality index data. The mathematical relationship of this neural network is as follows:

[0115] in, For the input layer neurons, corresponding to the fermentation process data (∫T) I 、∫I、∫T o 、∫RH、D、PR-PC); For intermediate layer neurons; z j The output layer neurons correspond to the predicted product quality index data (ANN, TNN).

[0116] 2) Train the neural network model.

[0117] First, construct the loss function: loss = (yp - y) 2

[0118] In the formula:

[0119] loss — the value of the loss function;

[0120] yp — the predicted value output by the neural network model, yp = z j ;

[0121] y—the actual value of the product quality indicator, which is the product quality indicator data.

[0122] Using preprocessed historical production data, a neural network model is trained using the gradient descent algorithm to minimize the sum of its losses across all training data. Based on the correlation coefficients between each fermentation process data point and the product quality index data calculated in step S103, an appropriate step size is selected for the gradient descent algorithm's weighting of the fermentation process data: 0.01 times the initial weight coefficient of the fermentation process data is used as the initial step size; the absolute values ​​of the correlation coefficients are arranged from smallest to largest, with the step size increased to 2-5 times for absolute values ​​below the median, and decreased to 0.1-0.5 times for absolute values ​​above the median; this improves training efficiency.

[0123] S105. Based on the real-time collected fermentation process data of the fermentation tank group, the trained neural network model is used to output the predicted product quality index data of the fermentation tank group; based on the product quality index data, with the goal of optimizing the optimal fermentation time of soy sauce and the overall production quality of the fermentation tank group, a dynamic management mathematical model is established, and a genetic algorithm is used to calculate the solution of the optimization goal, so as to realize the intelligent management of the fermentation tank group.

[0124] As shown in Figure 4, the genetic algorithm process for the dynamic management of the fermentation tank group is as follows:

[0125] 1) Establish a dynamic management mathematical model for the fermentation tank group.

[0126] The goal of dynamic management of fermenter groups is to optimize the overall product quality indicators of the fermenter groups. This is used to evaluate the fitness of individuals in the genetic algorithm, and the following mathematical model for the dynamic management of fermenter groups is established:

[0127] In the formula:

[0128] g1—ANN quality index and function of the fermenters participating in the genetic algorithm calculation of the fermenter group;

[0129] n—the number of fermenters in the fermenter group that participate in the genetic algorithm calculation;

[0130] ANN i —The ANN prediction value of the i-th fermenter; using the neural network model trained in step S104 and the real-time fermentation process data of the i-th fermenter, the ANN prediction value of the i-th fermenter is calculated.

[0131] g2—The TNN quality index and function of the fermenters participating in the genetic algorithm calculation of the fermenter group;

[0132] TNN i —The TNN prediction value of the i-th fermenter; using the neural network model trained in step S104 and the real-time fermentation process data of the i-th fermenter, the TNN prediction value of the i-th fermenter is calculated.

[0133] g3—The overall product quality index function used by the fermentation tank group in the genetic algorithm calculation;

[0134] The weighting coefficients of the total a-ANN predicted values;

[0135] The weighting coefficients of the total number of b-TNN predicted values.

[0136] 2) Chromosome encoding. The fermenters involved in the calculation are used as chromosomes, and the number of fermenters involved in the calculation is the number of chromosomes in the algorithm.

[0137] 3) Initialize and generate individual chromosomes to create the initial population.

[0138] 4) Decode the chromosomes, calculate the fitness of each individual, and evaluate the fitness value of each individual.

[0139] 5) Determine if the termination condition has been met. If it is, output the optimized calculation result; otherwise, go to 6).

[0140] 6) Select the next generation of population based on the selection strategy.

[0141] 7) Individuals undergo crossover operations according to the crossover probability.

[0142] 8) Individuals undergo mutation operations according to the mutation probability.

[0143] 9) A new generation of population is generated and returned to 4).

[0144] As shown in Figure 2, the method provided in this embodiment can be implemented through a fermentation tank group intelligent management system terminal and a fermentation online control system. The fermentation online control system is the main control system for fermentation tank group production. It collects production-related data from the fermentation tank group in real time and controls the valves, pumps, and motors of various actuators within the fermentation tank group to achieve fermentation tank production operation control. The fermentation tank group intelligent management system terminal is a terminal device that receives fermentation production data sent by the fermentation online control system through a communication port. All algorithm calculations for the various functional modules of the fermentation tank group intelligent management system can be completed on the terminal device. The terminal device then transmits the final optimization calculation results to the fermentation online control system, which completes the final fermentation tank production operation control.

[0145] Those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by a program instructing related hardware, and the corresponding program can be stored in a computer-readable storage medium.

[0146] It should be noted that although the method operations of the above embodiments are described in a specific order in the accompanying drawings, this does not require or imply that these operations must be performed in that specific order, or that all the operations shown must be performed to achieve the desired result. On the contrary, the order of execution of the described steps may be changed. Additionally or alternatively, certain steps may be omitted, multiple steps may be combined into one step, and / or one step may be broken down into multiple steps.

[0147] Example 2:

[0148] As shown in Figure 5, this embodiment provides an intelligent management system for a group of fermentation tanks based on the optimal fermentation time for soy sauce, including a data acquisition module 501, a data preprocessing module 502, a correlation coefficient calculation module 503, a model training module 504, and a dynamic management module 505, wherein:

[0149] The data acquisition module 501 is used to collect and store the production data of the fermentation tank group as historical production data of the fermentation tank group; the production data includes fermentation process data and product quality index data.

[0150] Data preprocessing module 502 is used to preprocess historical production data of the fermentation tank group;

[0151] The correlation coefficient calculation module 503 is used to calculate the correlation coefficient between fermentation process data and product quality indicators based on the pre-processed historical production data.

[0152] The model training module 504 is used to establish a neural network model of the mathematical relationship between fermentation process data and product quality index data based on preprocessed historical production data. The preprocessed historical production data is used as sample data, and the gradient descent algorithm is used to train the neural network model. When selecting the value of the gradient descent algorithm, the iteration calculation step size is selected according to the correlation coefficient between fermentation process data and product quality index data to improve calculation efficiency.

[0153] The dynamic management module 505 is used to output predicted product quality index data of the fermentation tank group based on the real-time collected fermentation process data of the fermentation tank group and the trained neural network model. Based on the product quality index data, with the goal of optimizing the optimal fermentation time of soy sauce and the overall production quality of the fermentation tank group, a dynamic management mathematical model is established, and the solution of the optimization goal is calculated by using a genetic algorithm to realize the intelligent management of the fermentation tank group.

[0154] The specific implementation of each module in this embodiment can be found in Embodiment 1 above, and will not be repeated here. It should be noted that the system provided in this embodiment is only illustrated by the division of the above functional modules. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure can be divided into different functional modules to complete all or part of the functions described above.

[0155] Example 3:

[0156] This embodiment provides a terminal device, which can be a computer, as shown in Figure 6. It is connected via a system bus 601 to a processor 602, a memory, an input device 603, a display 604, and a network interface 605. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium 606 and internal memory 607. The non-volatile storage medium 606 stores an operating system, computer programs, and a database. The internal memory 607 provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. When the processor 602 executes the computer programs stored in the memory, it implements the intelligent management method for fermentation tank groups based on the optimal fermentation time of soy sauce described in Embodiment 1, as follows:

[0157] Collect and store the production data of the fermentation tank group as historical production data of the fermentation tank group; the production data includes fermentation process data and product quality index data;

[0158] Preprocess the historical production data of the fermentation tank group;

[0159] Based on the pretreated historical production data, the correlation coefficient between fermentation process data and product quality indicators was calculated.

[0160] Based on preprocessed historical production data, a neural network model is established to establish the mathematical relationship between fermentation process data and product quality index data. The preprocessed historical production data is used as sample data, and the neural network model is trained using the gradient descent algorithm. When selecting the value of the gradient descent algorithm, the iteration calculation step size is selected according to the correlation coefficient between the fermentation process data and the product quality index data to improve the calculation efficiency.

[0161] Based on the real-time collected fermentation process data of the fermentation tank group, the trained neural network model is used to output the predicted product quality index data of the fermentation tank group. Based on the product quality index data, with the goal of optimizing the optimal fermentation time of soy sauce and the overall production quality of the fermentation tank group, a dynamic management mathematical model is established, and a genetic algorithm is used to calculate the solution of the optimization goal, so as to realize the intelligent management of the fermentation tank group.

[0162] Example 4:

[0163] This embodiment provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements the intelligent management method for fermentation tank groups based on the optimal fermentation time of soy sauce described in Embodiment 1 above, as follows:

[0164] Collect and store the production data of the fermentation tank group as historical production data of the fermentation tank group; the production data includes fermentation process data and product quality index data;

[0165] Preprocess the historical production data of the fermentation tank group;

[0166] Based on the pretreated historical production data, the correlation coefficient between fermentation process data and product quality indicators was calculated.

[0167] Based on preprocessed historical production data, a neural network model is established to establish the mathematical relationship between fermentation process data and product quality index data. The preprocessed historical production data is used as sample data, and the neural network model is trained using the gradient descent algorithm. When selecting the value of the gradient descent algorithm, the iteration calculation step size is selected according to the correlation coefficient between the fermentation process data and the product quality index data to improve the calculation efficiency.

[0168] Based on the real-time collected fermentation process data of the fermentation tank group, the trained neural network model is used to output the predicted product quality index data of the fermentation tank group. Based on the product quality index data, with the goal of optimizing the optimal fermentation time of soy sauce and the overall production quality of the fermentation tank group, a dynamic management mathematical model is established, and a genetic algorithm is used to calculate the solution of the optimization goal, so as to realize the intelligent management of the fermentation tank group.

[0169] It should be noted that the computer-readable storage medium in this embodiment can be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. The computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof.

[0170] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope disclosed in the present invention, based on the technical solution and inventive concept of the present invention, shall fall within the scope of protection of the present invention.

Claims

1. A method for intelligent management of fermentation tank groups based on the optimal fermentation time for soy sauce, characterized in that, The method includes: Collect and store the production data of the fermentation tank group as historical production data of the fermentation tank group; the production data includes fermentation process data and product quality index data; Preprocess the historical production data of the fermentation tank group; Based on the pretreated historical production data, the correlation coefficient between fermentation process data and product quality indicators was calculated. Based on preprocessed historical production data, a neural network model is established to establish the mathematical relationship between fermentation process data and product quality index data. The preprocessed historical production data is used as sample data, and the neural network model is trained using the gradient descent algorithm. When selecting the value of the gradient descent algorithm, the iteration calculation step size is selected according to the correlation coefficient between the fermentation process data and the product quality index data to improve the calculation efficiency. Based on the real-time collected fermentation process data of the fermentation tank group, the trained neural network model is used to output the predicted product quality index data of the fermentation tank group. Based on the product quality index data, with the goal of optimizing the optimal fermentation time of soy sauce and the overall production quality of the fermentation tank group, a dynamic management mathematical model is established, and a genetic algorithm is used to calculate the solution of the optimization goal, so as to realize the intelligent management of the fermentation tank group.

2. The intelligent management method for fermentation tank groups according to claim 1, characterized in that, The product quality index data includes the ammonia nitrogen content (ANN) and total nitrogen content (TNN) of soy sauce. The dynamic management mathematical model is as follows: In the formula, g1 and g2 are the ANN quality index and TNN quality index of the fermenters participating in the genetic algorithm calculation, respectively, n is the total number of fermenters participating in the genetic algorithm calculation, and ANN... i TNN i , respectively, represent the predicted ANN and TNN of the i-th fermenter; g3 is the overall product quality index of the fermenter group participating in the genetic algorithm calculation; a and b are the weight coefficients of the total predicted ANN and TNN, respectively.

3. The intelligent management method for fermentation tank groups according to any one of claims 1 and 2, characterized in that, In genetic algorithms, fermenters participating in the calculation are called chromosomes, and the total number of all fermenters participating in the calculation is the number of chromosomes.

4. The intelligent management method for fermentation tank groups according to claim 1, characterized in that, The step of using preprocessed historical production data as sample data and training a neural network model using the gradient descent algorithm includes: The fermentation cycle of soy sauce is divided into three independent fermentation stages: the initial fermentation stage, the middle fermentation stage, and the final fermentation stage. Each fermentation stage corresponds to a neural network model, that is, a neural network model is trained using historical production data of a fermentation stage. The three neural network models are constructed and trained using the same method. We will use one of the neural network models as an example: The input layer neurons are fermentation process data from preprocessed historical production data; The output layer neurons contain predicted product quality index data. The squared difference between the predicted product quality index data and the product quality index data in the preprocessed historical production data is used as the loss function, and the gradient descent algorithm is used to train the neural network model.

5. The intelligent management method for fermentation tank groups according to claim 4, characterized in that, The step of using real-time collected fermentation process data from the fermentation tank group and a trained neural network model to output predicted product quality index data for the fermentation tank group includes: Based on the real-time data collection time, the fermentation stage is determined, and then the corresponding trained neural network model is determined. After preprocessing the fermentation process data of the real-time collected fermentation tank group, the data is input into a defined neural network model, and the predicted product quality index data is output.

6. The intelligent management method for fermentation tank groups according to any one of claims 1 and 5, characterized in that, The fermentation process data includes the temperature inside the fermentation tank, light intensity, atmospheric temperature, atmospheric humidity, production date, and the row and column position of the fermentation tank in the tank area. The product quality index data includes the ammonia nitrogen content and total nitrogen content of the soy sauce. The preprocessing includes data cleaning and data standardization, wherein: Data cleaning includes: removing data where ammonia nitrogen is negatively correlated with time, as well as data points related to temperature, humidity, and light conditions that do not conform to natural laws; Data standardization processing includes: integrating the daily fermentation tank temperature, light intensity, atmospheric temperature, and atmospheric humidity for each fermenter, as well as classifying and quantifying the row and column positions for each fermenter.

7. The intelligent management method for fermentation tank groups according to any one of claims 1 and 4, characterized in that, The step of selecting the iterative calculation step size based on the correlation coefficient between fermentation process data and product quality index data includes: When determining the values ​​for the gradient descent algorithm, an appropriate step size should be selected for the step values ​​of the preprocessed fermentation process data: if the absolute value of the correlation coefficient is small, the step size should be increased appropriately; if the absolute value of the correlation coefficient is large, the step size should be decreased appropriately.

8. The intelligent management method for fermentation tank groups according to claim 1, characterized in that, The correlation coefficient was calculated using the Spearman rank correlation coefficient.

9. A smart management system for a group of fermentation tanks based on the optimal fermentation time for soy sauce, characterized in that, The system includes: The data acquisition module is used to collect and store the production data of the fermentation tank group as historical production data of the fermentation tank group; the production data includes fermentation process data and product quality index data. The data preprocessing module is used to preprocess the historical production data of the fermentation tank group; The correlation coefficient calculation module is used to calculate the correlation coefficient between fermentation process data and product quality indicators based on pre-processed historical production data. The model training module is used to establish a neural network model of the mathematical relationship between fermentation process data and product quality index data based on preprocessed historical production data. The preprocessed historical production data is used as sample data, and the gradient descent algorithm is used to train the neural network model. When selecting the value of the gradient descent algorithm, the iteration calculation step size is selected according to the correlation coefficient between fermentation process data and product quality index data to improve computational efficiency. The dynamic management module is used to output predicted product quality index data of the fermentation tank group based on the real-time collected fermentation process data of the fermentation tank group and a trained neural network model. Based on the product quality index data, with the goal of optimizing the optimal fermentation time of soy sauce and the overall production quality of the fermentation tank group, a dynamic management mathematical model is established, and the solution of the optimization objective is calculated by using a genetic algorithm to realize the intelligent management of the fermentation tank group.

10. A computer program product, characterized in that, It includes a computer program that, when executed by a processor, implements the intelligent management method for fermentation tank groups as described in any one of claims 1 to 8.

Images