A control method for preparing microbial inoculants based on the analysis of strain reproduction characteristics

Abstract

The present invention discloses a control method for preparing microbial inoculants based on the analysis of the reproduction characteristics of strains, aiming to improve the quality of inoculant preparation. By obtaining the reproduction data and metabolite data of the target strain, environmental sensitivity analysis is carried out to obtain the reproduction-environment sensitivity index and the metabolism-environment sensitivity index. Furthermore, a preparation quality prediction model is constructed to predict the quality of the target microbial inoculant in real time. By comparing the actual preparation quality data, it is judged whether there is metabolic imbalance, and the actual metabolite change data is obtained. These data are analyzed by Fourier transform to identify the metabolite change characteristics, and finally a metabolic optimization strategy is determined to optimize the preparation process. The present invention can effectively control the quality of microbial inoculants, improve production efficiency, and ensure the stability of products.

Description

Technical Field

[0001] The present invention relates to the technical field of microbial inoculant preparation, and particularly to a control method for microbial inoculant preparation based on the analysis of strain propagation characteristics. Background Art

[0002] With the continuous development of microbial technology, the application of microbial inoculants in multiple fields such as agriculture, environmental protection, and food industry has gradually received extensive attention. As an important biological agent, the preparation quality of microbial inoculants directly affects their application effects and economic benefits. Therefore, improving the production efficiency and quality of microbial inoculants has become an urgent problem to be solved in the current field of microbial preparation.

[0003] In the process of microbial inoculant preparation, the propagation characteristics of strains and the changes in metabolites are the key factors affecting the quality of the inoculant. Traditional methods for preparing microbial inoculants often rely on experience and experimental data, making it difficult to accurately predict the quality fluctuations during the preparation process and lacking systematic quality control means. This method is not only inefficient and costly, but also difficult to cope with complex production environments and the dynamic changes in strain metabolism.

[0004] Currently, there are some quality control methods based on strain growth characteristics or metabolic processes, but these methods generally lack consideration of environmental factors and the interactions between strains, and cannot achieve precise regulation of strain growth and metabolism under variable environmental conditions. Especially in large-scale industrial production, how to effectively control the quality of microbial inoculants and improve the controllability and stability of the preparation process remains a technical problem.

[0005] Therefore, the present invention provides a control method for microbial inoculant preparation based on the analysis of strain propagation characteristics, aiming to establish an efficient quality prediction and optimization mechanism by systematically analyzing the propagation characteristics and metabolite changes of target strains under different environmental conditions, thereby overcoming the defects in the prior art and improving the production quality and stability of microbial inoculants. Summary of the Invention

[0006] In order to solve at least one of the above technical problems, the present invention proposes a control method for microbial inoculant preparation based on the analysis of strain propagation characteristics.

[0007] The first aspect of the present invention provides a control method for microbial inoculant preparation based on the analysis of strain propagation characteristics, including:

[0008] Obtaining the data of the target strain types required for the preparation of the target microbial inoculant, obtaining the propagation data and metabolite data of each target strain under each environmental characteristic, performing environmental sensitivity analysis on the target strain according to the propagation data and metabolite data, and determining the propagation-environment sensitivity index and metabolism-environment sensitivity index of each target strain;

[0009] Construct a preparation quality prediction model for the target microbial agent according to the reproduction-environment sensitivity index and the metabolism-environment sensitivity index, and predict the current preparation quality of the target microbial agent according to the preparation quality prediction model to obtain a preparation quality prediction result;

[0010] Obtain the actual preparation quality data of the target microbial agent, compare the actual preparation quality data with the preparation quality prediction result, and judge the metabolic imbalance state of the target strain;

[0011] If the target strain shows a metabolic imbalance state, obtain the actual metabolite change data of the target microbial agent, and analyze the actual metabolite change data based on Fourier transform to determine the change characteristics of the actual metabolites;

[0012] Determine the metabolic optimization strategy for each time node according to the change characteristics of the actual metabolites.

[0013] In this solution, the method for obtaining the reproduction data and metabolite data of each target strain under various environmental characteristics, and performing environmental sensitivity analysis on the target strain according to the reproduction data and metabolite data to determine the reproduction-environment sensitivity index and the metabolism-environment sensitivity index of each target strain is specifically as follows:

[0014] Obtain the strain samples of each target strain type according to the target strain type data, place the strain samples under different environmental characteristics for reproduction and test preparation of the target microbial agent, and obtain the reproduction data and metabolite data of the target strain for reproduction and preparation of the target microbial agent under various environmental characteristics;

[0015] Construct a reproduction-environment data matrix and a metabolism-environment data matrix by combining the reproduction data and metabolite data with the respective environmental characteristics;

[0016] Calculate the covariance matrices of the reproduction-environment data matrix and the metabolism-environment data matrix, calculate the eigenvalues and eigenvectors of the covariance matrices, and select the eigenvectors corresponding to the top k largest eigenvalues to construct an eigenvector matrix;

[0017] Multiply the reproduction-environment data matrix and the metabolism-environment data matrix by the corresponding eigenvector matrix to obtain a reproduction-environment principal component score matrix and a metabolism-environment principal component score matrix;

[0018] Determine the influence degree of each environmental characteristic on the reproduction and metabolism of the target strain according to the reproduction-environment principal component score matrix and the metabolism-environment principal component score matrix, and obtain the reproduction-environment sensitivity index and the metabolism-environment sensitivity index.

[0019] In this solution, a preparation quality prediction model of the target microbial inoculant is constructed based on the reproduction-environment sensitivity index and the metabolism-environment sensitivity index. The current preparation quality of the target microbial inoculant is predicted according to the preparation quality prediction model to obtain a preparation quality prediction result, specifically as follows:

[0020] Obtain the reproduction situation data and metabolism situation data of the target strain during the preparation of the target microbial inoculant, perform a correlation analysis on the reproduction situation data and the metabolism situation data based on the Pearson correlation coefficient, determine the influence of the reproduction situation of the target strain on the metabolism situation, and obtain influence data;

[0021] Construct a preparation quality prediction model based on the fuzzy logic algorithm, and import the influence data into the preparation quality prediction model to construct the influence condition limit of the target strain reproduction on metabolism;

[0022] Take the preparation environment characteristics and the initial target strain number information for preparing the target microbial inoculant as the input data of the preparation quality prediction model;

[0023] Import the reproduction-environment sensitivity index and the metabolism-environment sensitivity index into the preparation quality prediction model to construct a reproduction quality membership function and a metabolism quality membership function of the target strain under different environment characteristics;

[0024] Construct a preparation quality fuzzy rule base for preparing the target microbial inoculant by the target strain under different strain numbers and different environment characteristics according to the influence data, the reproduction quality membership function and the metabolism quality membership function, and use the preparation quality fuzzy rule base as the prediction basis of the preparation quality prediction model;

[0025] Obtain the initial target strain number information and the current preparation environment characteristics of the target microbial inoculant preparation, import the initial target strain number information and the current preparation environment characteristics into the preparation quality prediction model, and determine the current reproduction quality and the current metabolism quality of the target strain under the current preparation environment characteristics according to the reproduction quality membership function, the metabolism quality membership function and the influence condition limit;

[0026] Compare the current reproduction quality and the current metabolism quality with the preparation quality fuzzy rule base, and predict the preparation quality of the target strain for preparing the target microbial inoculant under the current environment characteristics to obtain a preparation quality prediction result.

[0027] In this solution, obtain the actual preparation quality data of the target microbial inoculant, compare the actual preparation quality data with the preparation quality prediction result, and judge the metabolic imbalance state of the target strain, specifically as follows:

[0028] Take the predicted preparation quality result as the benchmark quality data for preparing the target microbial inoculant under the current environmental characteristics;

[0029] Obtain the actual preparation quality data of the target microbial inoculant prepared under the current environmental characteristics, compare the actual preparation quality data with the benchmark quality data, and calculate the Euclidean distance between the actual preparation quality data and the benchmark quality data;

[0030] Judge the metabolic imbalance state of the target strain according to the Euclidean distance.

[0031] In this solution, if the target strain shows a metabolic imbalance state, obtain the actual metabolite change data of the target microbial inoculant, and analyze the actual metabolite change data based on Fourier transform to determine the change characteristics of the actual metabolites. Specifically:

[0032] If the target strain shows a metabolic imbalance condition, monitor the metabolite changes of the target strain during the preparation of the target microbial inoculant based on a mass spectrometer to obtain the actual metabolite change data, and the actual metabolite change data includes the actual metabolite substance type change data and the actual metabolite concentration change data;

[0033] Perform a framing operation on the actual metabolite change data according to a preset window size based on a Hanning window to obtain actual metabolite change data segments;

[0034] Perform a fast Fourier transform on each actual metabolite change data segment, and calculate the spectral energy distribution of each actual metabolite change data segment;

[0035] Extract multi-scale features from the spectral energy distribution based on the wavelet packet decomposition algorithm. The multi-scale features include frequency components, amplitudes of frequency components, phases, and the proportion of band energy to obtain the change characteristics of the actual metabolites.

[0036] In this solution, determine the metabolic optimization strategy for each time node according to the change characteristics of the actual metabolites. Specifically:

[0037] Obtain the historical metabolite characteristics of different metabolic imbalance modes of the target strain, and construct a metabolic imbalance mode library by combining the metabolic imbalance modes with the corresponding historical metabolite characteristics;

[0038] Construct a bidirectional LSTM classification model based on the attention mechanism, perform a data linking operation on the metabolic imbalance mode library and the classification model, and construct a multi-scale feature vector from the change characteristics of the actual metabolites.

[0039] Import the multi-scale feature vectors into the classification model for dynamic weight matching with each historical metabolite feature in the metabolic imbalance pattern library, and output a confidence matrix for the metabolic imbalance types;

[0040] Determine the imbalance types of the target strain at each time node during the preparation of the target microbial inoculant according to the confidence matrix, where the imbalance types include metabolite inhibition and nutrient deficiency;

[0041] Determine the optimization strategy for each time node according to the imbalance type of the target strain at each time node.

[0042] In this solution, the step of determining the optimization strategy for each time node according to the imbalance type of the target strain at each time node is specifically as follows:

[0043] If the imbalance type is metabolite inhibition, determine the type of inhibitory metabolite that causes metabolic inhibition to the target strain according to the actual metabolite change data, analyze the actual metabolite change data based on the backpropagation network, determine the mutation point of the production rate of the inhibitory metabolite, and determine the metabolic change path of the inhibitory metabolite according to the mutation point of the production rate;

[0044] Determine the inhibition time point of the inhibitory metabolite on the metabolism of the target strain according to the metabolic change path, obtain the molecular size data of the inhibitory metabolite type, perform pore size matching of the nanoporous material according to the molecular size data to obtain the pore size distribution of the nanoporous material, and perform an adsorption operation on the inhibitory metabolite at the inhibition time point with the nanoporous material having the pore size distribution to obtain the first metabolic optimization strategy;

[0045] If the imbalance type is nutrient deficiency, obtain the change information of the reproduction quantity of each target strain during the preparation of the target microbial inoculant. If there is a situation where the reproduction quantity of one target strain continues to increase while the reproduction of another target strain stagnates, label the nutrient deficiency type as the microbial community nutrient competition type. If all target strains stop reproducing, label the nutrient deficiency type as the substrate component deficiency type;

[0046] When the nutrient deficiency type is the microbial community nutrient competition type, use the Lotka-Volterra competition model to invert the replacement trajectory of the dominant microbial community, identify the competitive strain type and competitive substrate type according to the replacement trajectory, determine the addition type of the competitive inhibitor and the substrate supplementation type according to the competitive strain type and competitive substrate type, and adjust the nutrient components for the preparation of the target microbial inoculant according to the addition type of the competitive inhibitor and the substrate supplementation type to obtain the second metabolic optimization strategy;

[0047] When the type of nutrient deficiency is substrate component deficiency, obtain the concentration change data of the substrate, determine the substrate deficiency type according to the concentration change data of the substrate, and supplement the nutrients of the target strain in the preparation process of the target microbial agent according to the substrate deficiency type to obtain the third metabolic optimization strategy.

[0048] The second aspect of the present invention also provides a control system for preparing microbial agents based on the analysis of strain reproduction characteristics. The system includes: a memory and a processor. The memory includes a program for the control method of preparing microbial agents based on the analysis of strain reproduction characteristics. When the program for the control method of preparing microbial agents based on the analysis of strain reproduction characteristics is executed by the processor, the following steps are realized:

[0049] Obtain the data of the target strain type required for the preparation of the target microbial agent, obtain the reproduction data and metabolite data of each target strain under each environmental characteristic, conduct environmental sensitivity analysis on the target strain according to the reproduction data and metabolite data, and determine the reproduction-environment sensitivity index and metabolism-environment sensitivity index of each target strain;

[0050] Construct a preparation quality prediction model for the target microbial agent according to the reproduction-environment sensitivity index and metabolism-environment sensitivity index, and predict the current preparation quality of the target microbial agent according to the preparation quality prediction model to obtain a preparation quality prediction result;

[0051] Obtain the actual preparation quality data of the target microbial agent, compare the actual preparation quality data with the preparation quality prediction result, and judge the metabolic imbalance state of the target strain;

[0052] If the target strain shows a metabolic imbalance state, obtain the actual metabolite change data of the target microbial agent, analyze the actual metabolite change data based on Fourier transform, and determine the change characteristics of the actual metabolite;

[0053] Determine the metabolic optimization strategy at each time node according to the change characteristics of the actual metabolite.

[0054] The present invention discloses a control method for preparing microbial inoculants based on the analysis of strain propagation characteristics, aiming to improve the quality of inoculant preparation. By obtaining the propagation data and metabolite data of the target strain, environmental sensitivity analysis is carried out to obtain the propagation-environment sensitivity index and the metabolism-environment sensitivity index, and then a preparation quality prediction model is constructed to predict the quality of the target microbial inoculant in real time. By comparing the actual preparation quality data, it is judged whether there is metabolic imbalance, and the actual metabolite change data are obtained. These data are analyzed by Fourier transform to identify the metabolite change characteristics, and finally a metabolic optimization strategy is determined to optimize the preparation process. The present invention can effectively control the quality of microbial inoculants, improve production efficiency, and ensure the stability of products. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 FIG. shows a flowchart of a control method for preparing microbial inoculants based on the analysis of strain propagation characteristics according to the present invention;

[0056] Figure 2 FIG. shows a flowchart of judging the metabolic imbalance state of the target strain according to the present invention;

[0057] Figure 3 FIG. shows a flowchart of determining the change characteristics of the actual metabolite according to the present invention;

[0058] Figure 4 FIG. shows a block diagram of a control system for preparing microbial inoculants based on the analysis of strain propagation characteristics according to the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0059] In order to more clearly understand the above objects, features and advantages of the present invention, the present invention will be further described in detail below with reference to the drawings and specific embodiments. It should be noted that, without conflict, the embodiments of the present application and the features in the embodiments can be combined with each other.

[0060] In the following description, many specific details are set forth in order to fully understand the present invention. However, the present invention may be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited by the specific embodiments disclosed below.

[0061] Figure 1 FIG. shows a flowchart of a control method for preparing microbial inoculants based on the analysis of strain propagation characteristics according to the present invention.

[0062] As Figure 1 shown, the first aspect of the present invention provides a control method for preparing microbial inoculants based on the analysis of strain propagation characteristics, including:

[0063] S102. Obtain the data of the types of target strains required for the preparation of the target microbial inoculant, obtain the reproduction data and metabolite data of each target strain under various environmental characteristics, perform environmental sensitivity analysis on the target strains according to the reproduction data and metabolite data, and determine the reproduction-environment sensitivity index and metabolism-environment sensitivity index of each target strain;

[0064] S104. Construct a preparation quality prediction model for the target microbial inoculant according to the reproduction-environment sensitivity index and metabolism-environment sensitivity index, predict the current preparation quality of the target microbial inoculant according to the preparation quality prediction model, and obtain a preparation quality prediction result;

[0065] S106. Obtain the actual preparation quality data of the target microbial inoculant, compare the actual preparation quality data with the preparation quality prediction result, and judge the metabolic imbalance state of the target strain;

[0066] S108. If the target strain shows a metabolic imbalance state, obtain the actual metabolite change data of the target microbial inoculant, analyze the actual metabolite change data based on Fourier transform, and determine the change characteristics of the actual metabolites;

[0067] S110. Determine the metabolic optimization strategy at each time node according to the change characteristics of the actual metabolites.

[0068] It should be noted that by obtaining the reproduction data and metabolite data of the target strain and performing environmental sensitivity analysis, it is possible to deeply understand the specific effects of different environmental characteristics on the growth and metabolism of the strain, thereby obtaining the reproduction-environment sensitivity index and metabolism-environment sensitivity index, providing a scientific basis for predicting the preparation quality of the inoculant. Secondly, the preparation quality prediction model established based on these indexes can predict the preparation quality of the target microbial inoculant in real time, discover potential quality fluctuations and risks in advance, and thus optimize the production process. In addition, through comparative analysis with the actual preparation quality data, it is possible to timely judge whether the strain shows a metabolic imbalance state, avoiding the influence of unstable factors on the quality of the inoculant. Once a metabolic imbalance is found, Fourier transform is used to analyze the change characteristics of metabolites, providing precise guidance for adjusting the strain metabolism process and optimizing production conditions to ensure the stability and high efficiency of the inoculant. Finally, by dynamically adjusting the optimization strategy, the preparation conditions of the microbial inoculant can be adjusted in real time at different production stages, thereby improving production efficiency and reducing costs. The types of target strains required for the preparation of the target microbial inoculant of the present invention include one or more, and the preparation of the target microbial inoculant first requires the reproduction of the target strain, and then the reproduction target strain is subjected to metabolic fermentation, and the metabolites can be used to prepare the target microbial inoculant, and finally the preparation product is obtained.

[0069] According to an embodiment of the present invention, obtaining the reproduction data and metabolite data of each target strain under various environmental characteristics, and performing environmental sensitivity analysis on the target strain according to the reproduction data and metabolite data to determine the reproduction-environment sensitivity index and metabolism-environment sensitivity index of each target strain, specifically:

[0070] Obtaining strain samples of each target strain type according to the target strain type data, placing the strain samples under different environmental characteristics for reproduction and preparation of the target microbial agent test, and obtaining the reproduction data and metabolite data of the target strain for reproduction and preparation of the target microbial agent under various environmental characteristics;

[0071] Constructing a reproduction-environment data matrix and a metabolism-environment data matrix with the reproduction data and metabolite data and each environmental characteristic;

[0072] Calculating the covariance matrix of the reproduction-environment data matrix and the metabolism-environment data matrix, calculating the eigenvalues and eigenvectors of the covariance matrix, and selecting the eigenvectors corresponding to the top k largest eigenvalues to construct an eigenvector matrix;

[0073] Multiplying the reproduction-environment data matrix and the metabolism-environment data matrix by the corresponding eigenvector matrix to obtain a reproduction-environment principal component score matrix and a metabolism-environment principal component score matrix;

[0074] Determining the influence degree of each environmental characteristic on the reproduction and metabolism of the target strain according to the reproduction-environment principal component score matrix and the metabolism-environment principal component score matrix, and obtaining the reproduction-environment sensitivity index and the metabolism-environment sensitivity index.

[0075] It should be noted that after constructing the reproduction-environment data matrix and metabolism-environment data matrix by introducing the principal component analysis algorithm (PCA) for the reproduction data, metabolite data, and each environmental characteristic, calculating the covariance matrix can reflect the linear correlation degree among the variables (environmental characteristics, reproduction, and metabolism data) in the dataset. The eigenvalue in the covariance matrix represents the data variance size in the direction of the corresponding eigenvector. The larger the variance, the more information the data in that direction contains. Multiplying the original data matrix (reproduction-environment data matrix and metabolism-environment data matrix) by the eigenvector matrix yields the principal component score matrix. The principal component score matrix maps the original reproduction-environment data matrix and metabolism-environment data matrix into a new space composed of the principal components (the directions determined by the eigenvectors corresponding to the first k largest eigenvalues). In this new space, each principal component represents a comprehensive influence pattern of an environmental factor combination on reproduction or metabolism. The magnitude of the principal component score reflects the performance degree of each sample (i.e., the strain experimental samples under different environmental characteristics) under these comprehensive influence patterns. The principal component with a large score change range corresponds to a more significant influence of the environmental factor combination on reproduction or metabolism. By analyzing the relationship between the principal component score and the original environmental characteristics, the influence degree of each environmental characteristic on reproduction and metabolism can be determined, and then the reproduction-environment sensitivity index and metabolism-environment sensitivity index can be obtained; the reproduction-environment sensitivity index reflects characteristics such as the growth rate and reproduction cycle of microorganisms under different environments; the metabolism-environment sensitivity index evaluates the change degree of metabolites (such as organic acids, amino acids, enzymes, etc.) of microbial strains under different environments; the reproduction data includes changes in reproduction speed, changes in the reproduction quantity of the target strain, and reproduction cycle data, and the metabolite data includes metabolite types and changes in metabolite content; the environmental characteristics include temperature, pH value, oxygen concentration, and nutrient content. The reproduction-environment sensitivity index reflects the significant degree of changes in the reproduction situation (such as growth rate, cell density, etc.) of the strain when environmental factors (such as temperature, nutrient content, etc.) change. The higher the index, the more sensitive the influence of this environmental characteristic on the reproduction of the strain, that is, a small change in environmental factors will cause a large change in the reproduction situation of the strain; the metabolism-environment sensitivity index reflects the significant degree of changes in the metabolites (species, concentration, yield, etc.) of the strain when environmental factors change. The higher the index, the more sensitive the influence of environmental factors on the metabolism of the strain, and the change in the environment will significantly affect the generation of metabolites.

[0076] According to an embodiment of the present invention, constructing a preparation quality prediction model for the target microbial inoculant based on the reproduction-environment sensitivity index and the metabolism-environment sensitivity index, and predicting the current preparation quality of the target microbial inoculant according to the preparation quality prediction model to obtain a preparation quality prediction result, specifically:

[0077] Obtain the data on the reproduction situation and metabolic situation of the target strain during the preparation of the target microbial inoculum, conduct a correlation analysis on the reproduction situation data and the metabolic situation data based on the Pearson correlation coefficient, determine the influence of the reproduction situation of the target strain on the metabolic situation, and obtain the influence data;

[0078] Construct a preparation quality prediction model based on the fuzzy logic algorithm, and import the influence data into the preparation quality prediction model to construct the influence condition limit of the reproduction of the target strain on metabolism;

[0079] Use the preparation environment characteristics and the initial target strain quantity information for preparing the target microbial inoculum as the input data of the preparation quality prediction model;

[0080] Import the reproduction-environment sensitivity index and the metabolism-environment sensitivity index into the preparation quality prediction model to construct the reproduction quality membership function and the metabolism quality membership function of the target strain under different environment characteristics;

[0081] Construct a preparation quality fuzzy rule base for preparing the target microbial inoculum by the target strain under different strain quantities and different environment characteristics according to the influence data, the reproduction quality membership function and the metabolism quality membership function, and use the preparation quality fuzzy rule base as the prediction basis of the preparation quality prediction model;

[0082] Obtain the initial target strain quantity information and the current preparation environment characteristics for preparing the target microbial inoculum, import the initial target strain quantity information and the current preparation environment characteristics into the preparation quality prediction model, and determine the current reproduction quality and the current metabolism quality of the target strain under the current preparation environment characteristics according to the reproduction quality membership function, the metabolism quality membership function and the influence condition limit;

[0083] Compare the current reproduction quality and the current metabolism quality with the preparation quality fuzzy rule base, and predict the preparation quality of preparing the target microbial inoculum by the target strain under the current environment characteristics to obtain the preparation quality prediction result.

[0084] It should be noted that since the reproduction situation of the target strain will affect the metabolic situation of the target strain during the subsequent preparation process of the target microbial agent, for example, if the reproduction quantity is too small, it will affect the production quantity of the subsequent metabolites. Therefore, by performing a correlation analysis on the reproduction situation and the metabolic situation, the influence of the reproduction situation of the target strain on the metabolic situation is determined; a preparation quality prediction model is constructed based on the fuzzy logic algorithm, and the above influence data is imported into the model to construct the influence condition restriction of the reproduction of the target strain on the metabolism. The growth and metabolism of microorganisms are greatly affected by environmental factors. Different environmental characteristics such as temperature, pH value, oxygen concentration, nutrient content, etc., as well as the initial strain quantity, will all have an impact on the preparation quality of the microbial agent. The fuzzy logic algorithm can well handle these multi-dimensional input information, comprehensively consider their influence on the preparation quality, and import the reproduction-environment sensitivity index and the metabolism-environment sensitivity index into the preparation quality prediction model to construct the reproduction quality membership function and the metabolism quality membership function of the target strain under different environmental characteristics. The membership function can fuzzify the reproduction and metabolism quality of the strain under different environments, describe the relationship between them and the environmental characteristics, which conforms to the actual situation that the quality state in the process of microbial growth and metabolism is not absolutely clear. Based on the influence data, the reproduction quality membership function and the metabolism quality membership function, a preparation quality fuzzy rule base for preparing the target microbial agent with different strain quantities and different environmental characteristics is constructed, and it is used as the prediction basis of the preparation quality prediction model. The fuzzy rule base integrates the relationships and influences between various factors, and contains fuzzy rules for the preparation quality in different situations, such as simulation rules that if the reproduction quality is high and the metabolism quality is high, then the preparation quality is high. The reproduction situation data includes the reproduction quantity and the reproduction cycle; the metabolism situation data includes the metabolite type, the metabolite content, and the metabolism rate; the influence data is the influence of different reproduction quantities and reproduction cycles of the target strain on the change of metabolite type, the change of content, and the change of metabolism rate; the influence condition restriction is the metabolic situation information of the target strain during the preparation process of the target microbial agent under different reproduction situations of the target strain, that is, the metabolic situation information of the target strain under different reproduction conditions; the higher the reproduction-environment sensitivity index in the reproduction membership function, the lower the reproduction quality under the corresponding environmental characteristics.

[0085] Figure 2 The flowchart for the present invention to judge the metabolic imbalance state of the target strain is shown.

[0086] According to an embodiment of the present invention, the actual preparation quality data of the target microbial agent is obtained, and the actual preparation quality data is compared with the preparation quality prediction result to judge the metabolic imbalance state of the target strain, specifically:

[0087] S202. Use the predicted preparation quality result as the reference quality data for preparing the target microbial inoculant under the current environmental characteristics.

[0088] S204. Obtain the actual preparation quality data of the target microbial inoculant under the current environmental characteristics, compare the actual preparation quality data with the reference quality data, and calculate the Euclidean distance between the actual preparation quality data and the reference quality data.

[0089] S206. Judge the metabolic imbalance state of the target strain according to the Euclidean distance.

[0090] It should be noted that if the Euclidean distance is greater than the preset value, it is considered that the target strain has metabolic imbalance. Since the reproduction and metabolism of the target strain are different under different environmental characteristics, the preparation quality of the target microbial inoculant will be different under different environmental characteristics. Therefore, using the predicted preparation quality result as the reference quality data for preparing the target microbial inoculant under the current environmental characteristics and comparing the reference quality data with the actual preparation quality data can find that when the Euclidean distance between the quality of preparing the target microbial inoculant under the current environmental characteristics and the actual preparation quality is greater than the preset value, it means that the difference between the actual preparation quality and the quality based on preparation is too large, and it can be considered that the metabolic imbalance of the target strain has led to the decline of the preparation quality of the target microbial inoculant.

[0091] Figure 3 The flowchart showing the change characteristics of the actual metabolites determined by the present invention is shown.

[0092] According to an embodiment of the present invention, if the target strain has a metabolic imbalance state, obtain the actual metabolite change data of the target microbial inoculant, and analyze the actual metabolite change data based on Fourier transform to determine the change characteristics of the actual metabolites. Specifically:

[0093] S302. If the target strain has a metabolic imbalance condition, monitor the metabolite changes of the target strain during the preparation of the target microbial inoculant based on a mass spectrometer to obtain actual metabolite change data, where the actual metabolite change data includes actual metabolite substance type change data and actual metabolite concentration change data.

[0094] S304. Based on the Hanning window, perform frame segmentation on the actual metabolite change data according to a preset window size to obtain actual metabolite change data segments.

[0095] S306. Perform a fast Fourier transform on each actual metabolite change data segment and calculate the spectral energy distribution of each actual metabolite change data segment.

[0096] S308, perform multi-scale feature extraction on the spectral energy distribution based on the wavelet packet decomposition algorithm. The multi-scale features include frequency components, amplitudes of frequency components, phases, and the proportion of band energy, to obtain the change characteristics of the actual metabolite.

[0097] It should be noted that the fast Fourier transform algorithm can convert the actual metabolite change data in the time domain to the frequency domain, making information such as periodicity and frequency characteristics that were difficult to discover in the time domain clearly presented. For example, through frequency domain analysis, the periodic frequency of the metabolite concentration change can be accurately known, which helps to understand the rhythm law of strain metabolism. On this basis, the wavelet packet decomposition algorithm further performs multi-scale feature extraction on the spectral energy distribution, and can obtain richer and more detailed metabolite change characteristics, such as frequency components, amplitudes of frequency components, phases, and the proportion of band energy. These characteristics can reflect the essence of metabolite changes from different angles. For example, the proportion of band energy can reflect the contribution degree of different frequency ranges in the overall metabolic change. Through real-time monitoring and analysis of the actual metabolite change data, the fast Fourier transform algorithm and the wavelet packet decomposition algorithm can sensitively capture the tiny anomalies in the metabolite changes. Before the target strain shows obvious metabolic imbalance, it may already be reflected in the change characteristics of the metabolite. For example, the amplitude of some frequency components shows abnormal fluctuations, or the proportion of band energy changes deviate from the normal range. These early abnormal signals can be used as warning indicators to help technicians timely discover potential metabolic imbalance problems.

[0098] According to the embodiments of the present invention, determining the metabolic optimization strategy for each time node according to the change characteristics of the actual metabolite is specifically as follows:

[0099] Obtain the historical metabolite characteristics of different metabolic imbalance modes of the target strain, and construct a metabolic imbalance mode library by associating the metabolic imbalance modes with the corresponding historical metabolite characteristics;

[0100] Construct a bidirectional LSTM classification model based on the attention mechanism, perform data linking operation on the metabolic imbalance mode library and the classification model, and construct a multi-scale feature vector from the change characteristics of the actual metabolite;

[0101] Import the multi-scale feature vector into the classification model to perform dynamic weight matching with each historical metabolite characteristic in the metabolic imbalance mode library, and output a confidence matrix of the metabolic imbalance type;

[0102] Determine the imbalance type of the target strain at each time node in the preparation process of the target microbial inoculant according to the confidence matrix. The imbalance types include metabolite inhibition and nutrient deficiency;

[0103] Determine the optimization strategy for each time node according to the imbalance type of the target strain at each time node.

[0104] It should be noted that by obtaining the historical metabolite characteristics of different metabolic imbalance patterns of the target strain and constructing a metabolic imbalance pattern library, and then combining with a bidirectional LSTM (Long Short-Term Memory Network) classification model based on the attention mechanism, the change characteristics of the actual metabolites can be compared and analyzed with the historical data. This method utilizes the empirical information in the historical data, making the analysis of the change characteristics of the current actual metabolites more comprehensive and accurate, so as to accurately judge the imbalance type of the target strain at each time node during the preparation process of the target microbial agent, such as metabolite inhibition or nutrient deficiency.

[0105] According to an embodiment of the present invention, the determining the optimization strategy for each time node according to the imbalance type of the target strain at each time node is specifically as follows:

[0106] If the imbalance type is metabolite inhibition, determine the type of inhibitory metabolite that causes metabolic inhibition to the target strain according to the actual metabolite change data, analyze the actual metabolite change data based on the backpropagation network, determine the mutation point of the production rate of the inhibitory metabolite, and determine the metabolic change path of the inhibitory metabolite according to the production rate mutation point;

[0107] Determine the inhibition time point of the inhibitory metabolite on the metabolism of the target strain according to the metabolic change path, obtain the molecular size data of the inhibitory metabolite type, perform pore size matching of the nanoporous material according to the molecular size data to obtain the pore size distribution of the nanoporous material, and perform an adsorption operation on the inhibitory metabolite at the inhibition time point with the nanoporous material with the pore size distribution to obtain the first metabolic optimization strategy;

[0108] It should be noted that during the preparation of microbial inoculants, target strains often face various imbalance problems, which affect the quality and preparation effect of the inoculants. Among them, metabolite inhibition and insufficient nutrient components are two relatively common types of problems. Metabolite inhibition may have a negative impact on strain metabolism due to the accumulation of certain metabolites, while insufficient nutrient components may hinder the reproduction of strains due to nutrient competition among bacteria or insufficient components of the substrate itself. The metabolic change path can reflect the entire dynamic process of the inhibitory metabolite from generation to accumulation. By analyzing each link and key node in the path, the changes of the inhibitory metabolite at different stages can be clearly seen. The production rate mutation point is an important indicator, which indicates that the production rate of the inhibitory metabolite has changed significantly after this point, possibly meaning that the inhibitory effect on the target strain metabolism is about to or has already started to increase. Analyzing along the metabolic change path starting from the production rate mutation point can accurately find the time point when the inhibitory metabolite substantially inhibits the target strain metabolism. Precise determination of the inhibition time point enables intervention at the most appropriate time during the adsorption operation of the nanoporous material, minimizing the impact of the inhibitory metabolite on the target strain metabolism, improving the metabolic efficiency and activity of the target strain, and ensuring the normal progress of the microbial inoculant preparation process. By precisely removing the inhibitory metabolite, it helps to maintain the normal growth and reproduction of the target strain, reduce the strain variation, death, etc. caused by metabolic inhibition, and thus improve the quality and stability of the microbial inoculant.

[0109] If the imbalance type is insufficient nutrient components, obtain the information on the change in the reproduction quantity of each target strain during the preparation of the target microbial inoculant. If there is a situation where the reproduction quantity of one target strain continues to increase while the reproduction of another target strain stagnates, label the insufficient nutrient component type as the bacterial colony nutrient competition type. If all target strains stop reproducing, label the insufficient nutrient component type as the substrate component insufficiency type.

[0110] When the insufficient nutrient component type is the bacterial colony nutrient competition type, use the Lotka-Volterra competition model to invert the replacement trajectory of the dominant bacterial colonies, identify the competing strain type and the competing substrate type according to the replacement trajectory, determine the addition type of the competitive inhibitor and the substrate supplement type according to the competing strain type and the competing substrate type, and adjust the nutrient components for the preparation of the target microbial inoculant according to the addition type of the competitive inhibitor and the substrate supplement type to obtain the second metabolic optimization strategy.

[0111] It should be noted that by analyzing the information on the change in the reproduction quantity of each target strain during the preparation of the target microbial inoculant, it is possible to accurately distinguish whether the insufficient nutrient components are caused by nutrient competition among the microbial communities or insufficient substrate components; the Lotka-Volterra competition model is a classic ecological model used to describe the competition relationship among two or more species in an environment with limited resources. In the environment of microbial inoculant preparation, different target strains can be regarded as different competing species, and they compete for survival conditions such as limited nutrient resources. Based on the logistic equation of population growth, this model takes into account factors such as the growth rate of each population itself, the environmental carrying capacity, and the competition coefficients among different populations, and can quantitatively describe the change relationship of the quantities of different populations over time from a mathematical perspective. The replacement trajectory of the dominant microbial community refers to the process and path of the change in the dominant position of different target strains in the microbial community due to factors such as nutrient competition over time during the preparation of the microbial inoculant. Based on the accurate identification of the types of competing strains and the types of competing substrates, the types of competitive inhibitors to be added and the types of substrate supplements are determined, and then the nutrient components for the preparation of the target microbial inoculant are reasonably adjusted. This can effectively regulate the competition relationship among the microbial communities, inhibit excessive competition, promote the balanced growth of each strain, improve the quality and stability of the microbial inoculant, ensure that all strains in the microbial inoculant can play their due roles, and enhance the overall performance of the inoculant.

[0112] When the type of insufficient nutrient components is the type of insufficient substrate components, obtain the concentration change data of the substrate, determine the type of substrate deficiency according to the concentration change data of the substrate, and supplement the nutrient components of the target strain during the preparation of the target microbial inoculant according to the type of substrate deficiency to obtain the third metabolic optimization strategy.

[0113] It should be noted that by obtaining the concentration change data of the substrate to determine the type of substrate deficiency, it is possible to accurately identify which substrate or substrates are insufficient during the preparation of the target microbial inoculant. This avoids the blind supplementation of nutrient components that may be caused by the inability to accurately judge the type of substrate deficiency; supplementing the nutrient components of the target strain based on the determined type of substrate deficiency can achieve targeted and accurate supplementation. This makes the supplementation of nutrient components match the actual needs of the strain, neither failing to solve the problem due to insufficient supplementation nor causing waste of resources due to excessive supplementation, which helps to improve the utilization efficiency of nutrient components, ensure that the target strain can obtain a suitable nutrient environment for its growth and reproduction, promote the normal growth and metabolism of the strain, and ultimately improve the preparation quality of the target microbial inoculant.

[0114] According to the embodiments of the present invention, it further includes:

[0115] Obtain the data of the change in the number of microorganisms attached to the target microorganism inoculant preparation equipment within a preset time period, and construct the time-varying sequence data of the microorganism attachment of the preparation equipment according to the data of the change in the number of microorganisms attached;

[0116] Construct a prediction model for the change in microorganism attachment based on the long short-term memory network. Divide the time-varying sequence data of the microorganism attachment into prediction set data and training set data according to a preset ratio, and import the training set data into the prediction model for the change in microorganism attachment for training operations;

[0117] Import the prediction set data into the trained prediction model for the change in microorganism attachment to predict the change in microorganism attachment within a preset future time period, and obtain the prediction result of the change in the number of microorganisms attached;

[0118] Determine the degree of contamination of the target microorganism inoculant preparation by the preparation equipment within a preset future time period according to the prediction result of the change in the number of microorganisms attached;

[0119] Obtain the preparation purity requirement of the target microorganism inoculant, and determine the cleaning cycle of the preparation equipment according to the preparation purity requirement and the degree of contamination.

[0120] It should be noted that during the continuous production process of microorganism inoculants, the production equipment operates for a long time and ages, and the phenomenon of microorganism attachment in the pipeline becomes increasingly serious. Traditional cleaning methods are often based on fixed time intervals or empirical judgments, lacking scientific basis. At the same time, due to the complex and changeable situation of microorganism attachment, it is difficult to accurately estimate when it will have a serious impact on the preparation of target microorganism inoculants, resulting in either premature equipment cleaning causing waste of resources or too late cleaning causing the quality of microorganism inoculants to be contaminated, thereby affecting product quality and production efficiency; by obtaining the data of the change in the number of microorganisms attached and constructing the time-varying sequence data, and using the long short-term memory network to construct a prediction model, it is possible to effectively predict the future change in microorganism attachment. Based on the prediction result of the change in the number of microorganisms attached, it is possible to accurately determine the degree of contamination of the preparation equipment for the target microorganism inoculant preparation within a preset future time period, enabling production personnel to clearly understand the contamination status of the equipment and make preparations in advance. Combining the preparation purity requirement of the target microorganism inoculant and the degree of equipment contamination, it is possible to scientifically and reasonably determine the cleaning cycle of the preparation equipment. This not only avoids the waste of resources and equipment loss caused by excessive cleaning, but also prevents the decline in the quality of the inoculant due to untimely cleaning, thereby ensuring product quality while reducing production costs and improving production efficiency.

[0121] Figure 4 The block diagram of a control system for preparing microorganism inoculants based on the analysis of strain reproduction characteristics according to the present invention is shown.

[0122] In a second aspect of the present invention, there is also provided a control system 4 for preparing a microbial inoculant based on the analysis of the reproduction characteristics of strains, the system comprising: a memory 41 and a processor 42. The memory includes a program for a method for controlling the preparation of a microbial inoculant based on the analysis of the reproduction characteristics of strains. When the program for the method for controlling the preparation of a microbial inoculant based on the analysis of the reproduction characteristics of strains is executed by the processor, the following steps are implemented:

[0123] Obtain the data of the target strain type required for the preparation of the target microbial inoculant, obtain the reproduction data and metabolite data of each target strain under various environmental characteristics, perform environmental sensitivity analysis on the target strain according to the reproduction data and metabolite data, and determine the reproduction-environment sensitivity index and the metabolism-environment sensitivity index of each target strain;

[0124] Construct a preparation quality prediction model for the target microbial inoculant according to the reproduction-environment sensitivity index and the metabolism-environment sensitivity index, predict the current preparation quality of the target microbial inoculant according to the preparation quality prediction model, and obtain a preparation quality prediction result;

[0125] Obtain the actual preparation quality data of the target microbial inoculant, compare the actual preparation quality data with the preparation quality prediction result, and judge the metabolic imbalance state of the target strain;

[0126] If the target strain shows a metabolic imbalance state, obtain the actual metabolite change data of the target microbial inoculant, analyze the actual metabolite change data based on Fourier transform, and determine the change characteristics of the actual metabolite;

[0127] Determine the metabolic optimization strategy for each time node according to the change characteristics of the actual metabolite.

[0128] The present invention discloses a method for controlling the preparation of a microbial inoculant based on the analysis of the reproduction characteristics of strains, aiming to improve the quality of inoculant preparation. By obtaining the reproduction data and metabolite data of the target strain, performing environmental sensitivity analysis, obtaining the reproduction-environment sensitivity index and the metabolism-environment sensitivity index, and then constructing a preparation quality prediction model to predict the quality of the target microbial inoculant in real time. By comparing the actual preparation quality data, it is judged whether there is a metabolic imbalance, and the actual metabolite change data is obtained. These data are analyzed using Fourier transform to identify the change characteristics of metabolites, and finally the metabolic optimization strategy is determined to optimize the preparation process. The present invention can effectively control the quality of microbial inoculants, improve production efficiency, and ensure the stability of products.

[0129] In several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed with each other can be through some interfaces. The indirect coupling or communication connection of devices or units can be electrical, mechanical, or other forms.

[0130] The units described above as separate components may or may not be physically separated. The components shown as units may or may not be physical units. They can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0131] In addition, each functional unit in the embodiments of the present invention can be all integrated in a processing unit, or each unit can be separately used as a unit, or two or more units can be integrated in one unit. The above-mentioned integrated units can be implemented in the form of hardware, or in the form of a combination of hardware and software functional units.

[0132] Those of ordinary skill in the art can understand that all or part of the steps to implement the above method embodiments can be completed by hardware related to program instructions. The foregoing program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps including the above method embodiments. The foregoing storage medium includes various media that can store program codes, such as removable storage devices, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disks, or optical discs.

[0133] Alternatively, if the above-mentioned integrated units of the present invention are implemented in the form of software functional modules and sold or used as independent products, they can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiments of the present invention, in essence, or the part that contributes to the prior art can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the methods described in the various embodiments of the present invention. The foregoing storage medium includes various media that can store program codes, such as removable storage devices, ROM, RAM, magnetic disks, or optical discs.

[0134] As described above, it is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of changes or substitutions, which should all be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claimed rights.

Claims

1. A method for controlling the preparation of a microbial inoculant based on the analysis of the reproduction characteristics of strains, characterized in that, It includes the following steps: Obtain the data of the target strain types required for the preparation of the target microbial inoculant, obtain the reproduction data and metabolite data of each target strain under various environmental characteristics, perform environmental sensitivity analysis on the target strain according to the reproduction data and metabolite data, and determine the reproduction-environment sensitivity index and metabolism-environment sensitivity index of each target strain; Construct a preparation quality prediction model for the target microbial inoculant according to the reproduction-environment sensitivity index and metabolism-environment sensitivity index, predict the current preparation quality of the target microbial inoculant according to the preparation quality prediction model, and obtain a preparation quality prediction result; Obtain the actual preparation quality data of the target microbial inoculant, compare the actual preparation quality data with the preparation quality prediction result, and judge the metabolic imbalance state of the target strain; If the target strain shows a metabolic imbalance state, obtain the actual metabolite change data of the target microbial inoculant, analyze the actual metabolite change data based on Fourier transform, and determine the change characteristics of the actual metabolites; Determine the metabolic optimization strategy at each time node according to the change characteristics of the actual metabolites; The determining the metabolic optimization strategy at each time node according to the change characteristics of the actual metabolites is specifically: Obtain the historical metabolite characteristics of different metabolic imbalance modes of the target strain, and construct a metabolic imbalance mode library by combining the metabolic imbalance modes with the corresponding historical metabolite characteristics; Construct a bidirectional LSTM classification model based on the attention mechanism, perform a data linking operation on the metabolic imbalance mode library and the classification model, and construct a multi-scale feature vector from the change characteristics of the actual metabolites; Import the multi-scale feature vector into the classification model to perform dynamic weight matching with each historical metabolite characteristic in the metabolic imbalance mode library, and output a confidence matrix of the metabolic imbalance type; Determine the imbalance type of the target strain at each time node during the preparation of the target microbial inoculant according to the confidence matrix, and the imbalance types include metabolite inhibition and nutrient deficiency; Determine the optimization strategy at each time node according to the imbalance type of the target strain at each time node; The determining the optimization strategy at each time node according to the imbalance type of the target strain at each time node is specifically: If the imbalance type is metabolite inhibition, determine the type of inhibitory metabolite that causes metabolic inhibition to the target strain according to the actual metabolite change data, analyze the actual metabolite change data based on the backpropagation network, determine the production rate mutation point of the inhibitory metabolite, and determine the metabolic change path of the inhibitory metabolite according to the production rate mutation point; Determine the inhibition time point of the inhibitory metabolite on the metabolism of the target strain according to the metabolic change path, obtain the molecular size data of the inhibitory metabolite type, perform pore size matching of the nanoporous material according to the molecular size data to obtain the pore size distribution of the nanoporous material, and perform an adsorption operation on the inhibitory metabolite at the inhibition time point with the nanoporous material with the pore size distribution to obtain the first metabolic optimization strategy; If the imbalance type is insufficient nutrient components, obtain the information on the change in the reproduction quantity of each target strain during the preparation of the target microbial inoculum. If there is a situation where the reproduction quantity of a certain target strain continuously increases while the reproduction of another target strain stagnates, label the insufficient nutrient component type as the microbial community nutrient competition type. If all target strains stop reproducing, label the insufficient nutrient component type as the substrate component deficiency type; When the insufficient nutrient component type is the microbial community nutrient competition type, use the Lotka-Volterra competition model to invert the replacement trajectory of the dominant microbial community, identify the competing strain type and the competing substrate type based on the replacement trajectory, determine the addition type of the competitive inhibitor and the substrate supplementation type based on the competing strain type and the competing substrate type, and adjust the nutrient components for the preparation of the target microbial inoculum according to the addition type of the competitive inhibitor and the substrate supplementation type to obtain the second metabolic optimization strategy; When the insufficient nutrient component type is the substrate component deficiency type, obtain the concentration change data of the substrate, determine the substrate deficiency type based on the concentration change data of the substrate, and supplement the nutrient components of the target strain during the preparation of the target microbial inoculum according to the substrate deficiency type to obtain the third metabolic optimization strategy.

2. The preparation control method of a microbial inoculant based on the analysis of strain propagation characteristics according to claim 1, wherein, The obtaining of the reproduction data and metabolite data of each target strain under each environmental characteristic, and the environmental sensitivity analysis of the target strain based on the reproduction data and metabolite data to determine the reproduction-environment sensitivity index and the metabolism-environment sensitivity index of each target strain are specifically as follows: Obtain the strain samples of each target strain type according to the target strain type data, place the strain samples under different environmental characteristics for reproduction and the experimental preparation of the target microbial inoculum, and obtain the reproduction data and metabolite data of the target strain for reproduction and the preparation of the target microbial inoculum under each environmental characteristic; Construct a reproduction-environment data matrix and a metabolism-environment data matrix with the reproduction data and metabolite data and each environmental characteristic; Calculate the covariance matrix of the reproduction-environment data matrix and the metabolism-environment data matrix, calculate the eigenvalues and eigenvectors of the covariance matrix, and select the eigenvectors corresponding to the top k largest eigenvalues to construct an eigenvector matrix; Multiply the reproduction-environment data matrix and the metabolism-environment data matrix by the corresponding eigenvector matrix to obtain a reproduction-environment principal component score matrix and a metabolism-environment principal component score matrix; Determine the influence degree of each environmental characteristic on the reproduction and metabolism of the target strain based on the reproduction-environment principal component score matrix and the metabolism-environment principal component score matrix to obtain the reproduction-environment sensitivity index and the metabolism-environment sensitivity index.

3. A method for controlling the preparation of a microbial inoculant based on the analysis of the reproductive characteristics of strains, characterized in that, The construction of the preparation quality prediction model of the target microbial inoculum based on the reproduction-environment sensitivity index and the metabolism-environment sensitivity index, and the prediction of the current preparation quality of the target microbial inoculum according to the preparation quality prediction model to obtain the preparation quality prediction result are specifically as follows: Obtain the data on the reproduction situation and metabolic situation of the target strain during the preparation of the target microbial inoculum, perform a correlation analysis on the reproduction situation data and the metabolic situation data based on the Pearson correlation coefficient, determine the impact of the reproduction situation of the target strain on the metabolic situation, and obtain the impact data; Construct a preparation quality prediction model based on the fuzzy logic algorithm, and import the impact data into the preparation quality prediction model to construct the impact condition limit of the reproduction of the target strain on metabolism; Use the preparation environment characteristics and the information on the initial number of target strains for preparing the target microbial inoculum as the input data of the preparation quality prediction model; Import the reproduction-environment sensitivity index and the metabolism-environment sensitivity index into the preparation quality prediction model to construct the reproduction quality membership function and the metabolism quality membership function of the target strain under different environment characteristics; Construct a preparation quality fuzzy rule base for preparing the target microbial inoculum by the target strain under different strain numbers and different environment characteristics according to the impact data, the reproduction quality membership function and the metabolism quality membership function, and use the preparation quality fuzzy rule base as the prediction basis of the preparation quality prediction model; Obtain the information on the initial number of target strains for preparing the target microbial inoculum and the current preparation environment characteristics, import the information on the initial number of target strains and the current preparation environment characteristics into the preparation quality prediction model, and determine the current reproduction quality and the current metabolism quality of the target strain under the current preparation environment characteristics according to the reproduction quality membership function, the metabolism quality membership function and the impact condition limit; Compare the current reproduction quality and the current metabolism quality with the preparation quality fuzzy rule base, and predict the preparation quality of the target strain for preparing the target microbial inoculum under the current environment characteristics to obtain the preparation quality prediction result.

4. A method for controlling the preparation of a microbial inoculant based on the analysis of the reproductive characteristics of strains, characterized in that, Obtain the actual preparation quality data of the target microbial inoculum, compare the actual preparation quality data with the preparation quality prediction result, and judge the metabolic imbalance state of the target strain, specifically: Use the preparation quality prediction result as the reference quality data for preparing the target microbial inoculum under the current environment characteristics; Obtain the actual preparation quality data for preparing the target microbial inoculum under the current environment characteristics, compare the actual preparation quality data with the reference quality data, and calculate the Euclidean distance between the actual preparation quality data and the reference quality data; Judge the metabolic imbalance state of the target strain according to the Euclidean distance.

5. A method for controlling the preparation of a microbial inoculant based on the analysis of the reproduction characteristics of strains, characterized in that, If the target strain shows a metabolic imbalance state, obtain the actual metabolic product change data of the target microbial inoculum, and analyze the actual metabolic product change data based on Fourier transform to determine the change characteristics of the actual metabolic products, specifically: If the target strain shows a metabolic imbalance condition, monitor the change of the metabolic products of the target strain during the preparation of the target microbial inoculum based on a mass spectrometer to obtain the actual metabolic product change data, and the actual metabolic product change data includes the actual metabolic product substance type change data and the actual metabolic product concentration change data; Frame the actual metabolite change data based on the Hanning window according to a preset window size to obtain actual metabolite change data segments; Perform fast Fourier transform on each of the actual metabolite change data segments and calculate the spectral energy distribution of each actual metabolite change data segment; Extract multi-scale features from the spectral energy distribution based on the wavelet packet decomposition algorithm. The multi-scale features include frequency components, amplitudes of frequency components, phases, and the proportion of band energy, to obtain the change characteristics of the actual metabolites.

6. A preparation control system for microbial inoculants based on the analysis of the reproductive characteristics of strains, characterized in that, The microbial inoculant preparation control system based on the analysis of strain reproduction characteristics includes a memory and a processor. The memory includes a program for the microbial inoculant preparation control method based on the analysis of strain reproduction characteristics. When the program for the microbial inoculant preparation control method based on the analysis of strain reproduction characteristics is executed by the processor, the following steps are implemented: Obtain the target strain type data required for the preparation of the target microbial inoculant, obtain the reproduction data and metabolite data of each target strain under various environmental characteristics, and perform environmental sensitivity analysis on the target strain according to the reproduction data and metabolite data to determine the reproduction-environment sensitivity index and metabolism-environment sensitivity index of each target strain; Construct a preparation quality prediction model for the target microbial inoculant according to the reproduction-environment sensitivity index and metabolism-environment sensitivity index, and predict the current preparation quality of the target microbial inoculant according to the preparation quality prediction model to obtain a preparation quality prediction result; Obtain the actual preparation quality data of the target microbial inoculant, compare the actual preparation quality data with the preparation quality prediction result, and judge the metabolic imbalance state of the target strain; If the target strain shows a metabolic imbalance state, obtain the actual metabolite change data of the target microbial inoculant, and analyze the actual metabolite change data based on Fourier transform to determine the change characteristics of the actual metabolites; Determine the metabolic optimization strategy for each time node according to the change characteristics of the actual metabolites; The determining the metabolic optimization strategy for each time node according to the change characteristics of the actual metabolites is specifically: Obtain the historical metabolite characteristics of different metabolic imbalance modes of the target strain, and construct a metabolic imbalance mode library by associating the metabolic imbalance modes with the corresponding historical metabolite characteristics; Construct a bidirectional LSTM classification model based on the attention mechanism, perform a data linking operation between the metabolic imbalance mode library and the classification model, and construct a multi-scale feature vector from the change characteristics of the actual metabolites; Import the multi-scale feature vector into the classification model to perform dynamic weight matching with each historical metabolite characteristic in the metabolic imbalance mode library, and output a confidence matrix of the metabolic imbalance type; Determine the imbalance type of the target strain at each time node during the preparation of the target microbial inoculant according to the confidence matrix. The imbalance types include metabolite inhibition and nutrient deficiency; Determine the optimization strategy for each time node according to the imbalance type of the target strain at each time node; Determining the optimization strategy for each time node according to the imbalance type of the target strain at each time node specifically includes: If the imbalance type is metabolite inhibition, determining the type of inhibitory metabolite that causes metabolic inhibition to the target strain according to the actual metabolite change data, analyzing the actual metabolite change data based on the backpropagation network, determining the production rate mutation point of the inhibitory metabolite, and determining the metabolic change path of the inhibitory metabolite according to the production rate mutation point; Determining the inhibition time point of the inhibitory metabolite on the metabolism of the target strain according to the metabolic change path, obtaining the molecular size data of the inhibitory metabolite type, performing pore size matching of the nanoporous material according to the molecular size data to obtain the pore size distribution of the nanoporous material, and performing an adsorption operation on the inhibitory metabolite at the inhibition time point according to the nanoporous material with the pore size distribution to obtain the first metabolic optimization strategy; If the imbalance type is insufficient nutrient components, obtaining the change information of the reproduction quantity of each target strain during the preparation of the target microbial agent. If there is a situation where the reproduction quantity of a certain target strain continues to increase while the reproduction of another target strain stagnates, the insufficient nutrient component type is labeled as the microbial community nutrient competition type. If all target strains stop reproducing, the insufficient nutrient component type is labeled as the substrate component insufficiency type; When the insufficient nutrient component type is the microbial community nutrient competition type, using the Lotka-Volterra competition model to invert the replacement trajectory of the dominant microbial community, identifying the competing strain type and the competing substrate type according to the replacement trajectory, determining the addition type of the competition inhibitor and the substrate supplement type according to the competing strain type and the competing substrate type, and adjusting the nutrient components for the preparation of the target microbial agent according to the addition type of the competition inhibitor and the substrate supplement type to obtain the second metabolic optimization strategy; When the insufficient nutrient component type is the substrate component insufficiency type, obtaining the concentration change data of the substrate, determining the substrate deficiency type according to the concentration change data of the substrate, and supplementing the nutrient components of the target strain during the preparation process of the target microbial agent according to the substrate deficiency type to obtain the third metabolic optimization strategy.

Images