Pear fire blight resistant pear stock tissue culture environment control method
By applying machine learning models to carry out intelligent environmental control in plant tissue culture, the problem of lack of scientificity and intelligence in traditional methods is solved, and more efficient and stable tissue culture is achieved, and the plant's disease resistance is enhanced.
Patent Information
- Application Number
- CN202510189034.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-02-20
- Publication Date
- 2025-06-10
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Traditional plant tissue culture methods lack scientific quantitative indicators and intelligent decision-making tools in environmental control, which makes it difficult to ensure the cultivation efficiency and stability, especially in pear rootstock tissue culture that is resistant to pear fire blight.
Pre-trained machine learning model is adopted, combined with classification model and regression model, and real-time monitoring of plant tissue metabolism and growth data, scientifically judge the necessity and timing of medium replacement, and dynamically adjust the medium composition, and set the buffer stage to reduce the sensitivity of plants to medium mutations.
Through intelligent management of the culture medium replacement process, the success rate of tissue culture and the disease resistance of plants are improved, the healthy and stable plant growth is ensured, and the risks of resource waste and growth stress are reduced.
Smart Images
Figure CN120123683A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of plant tissue culture, and more specifically, to a method for controlling the tissue culture environment of a pear rootstock resistant to fire blight of pear. Background Art
[0002] Fire blight of pear is a bacterial disease that causes serious harm to pear tree planting. Its pathogen is Erwinia amylovora, which mainly infects rosaceous plants such as pears and apples. The bacterium spreads through flowers, leaves, branches and fruits, and rapidly spreads under suitable temperature and humidity conditions, resulting in serious economic losses. Controlling fire blight of pear is difficult. Traditional control methods such as spraying chemical agents and removing diseased trees are costly and have limited effects, and may also have negative impacts on the environment and ecology.
[0003] In modern agricultural production, improving the disease resistance of pear rootstocks has become an important way to solve the problem of fire blight of pear. Tissue culture technology is a method of culturing plant cells, tissues or organs under sterile conditions, which can be used for rapid propagation of disease-resistant rootstocks. However, direct replacement of the culture medium at different stages is likely to cause plant growth stress, resulting in metabolic disorders, growth stagnation or even culture failure. Moreover, the current control of the culture environment is still mainly based on experience, lacking scientific quantitative indicators and intelligent decision-making tools, and it is difficult to achieve efficient and stable tissue culture. Therefore, the present invention proposes a method for controlling the tissue culture environment of a pear rootstock resistant to fire blight of pear in order to solve the above problems. Summary of the Invention
[0004] To achieve the above object, the present invention provides the following technical solutions:
[0005] A method for controlling the tissue culture environment of a pear rootstock resistant to fire blight of pear, comprising the following steps: during the tissue culture of a pear rootstock resistant to fire blight of pear, obtaining the current culture medium usage data set caused by plant tissue metabolism and the plant tissue growth data set;
[0006] Substituting the current culture medium usage data set and the plant tissue growth data set into a pre-trained machine learning model together, outputting the result of whether to replace the culture medium, and when it is necessary to replace the culture medium, predicting the components of the culture medium to obtain the parameter set two corresponding to the optimized culture medium;
[0007] Performing feature extraction and analysis on the parameter set two corresponding to the optimized culture medium and the parameter set one corresponding to the current culture medium to obtain an analysis result;
[0008] Based on the analysis result, performing a buffer determination to determine the duration of the buffer stage to be experienced before replacing the current culture medium with the optimized culture medium corresponding to the parameter set two and execute it.
[0009] In a preferred embodiment, the pre-trained machine learning model is a joint model, which is composed of a pre-trained classification model and a regression model.
[0010] In a preferred embodiment, the classification model is used to determine whether the current culture medium needs to be replaced. When the classification model outputs a result that the current culture medium needs to be replaced, the prediction mechanism is triggered, the regression model is enabled, and the components of the culture medium to be replaced are predicted to obtain the second set of parameters corresponding to the optimized culture medium.
[0011] In a preferred embodiment, the feature extraction and analysis of the second set of parameters corresponding to the optimized culture medium and the first set of parameters corresponding to the current culture medium refer to:
[0012] Based on the second set of parameters corresponding to the optimized culture medium and the first set of parameters corresponding to the current culture medium, a data set of culture medium component difference features and a data set of plant applicability deviation features are respectively extracted. Then, the difference analysis is performed on the data set of culture medium component difference features to generate a culture medium deviation index, and the deviation analysis is performed on the data set of plant applicability deviation features to generate a plant adaptability index.
[0013] In a preferred embodiment, the acquisition logic of the culture medium deviation index is as follows:
[0014] There are n culture medium factors in the data set of culture medium component difference features, and the deviation of the i-th factor is ΔX i , and the calculation formula is:
[0015] X act，i represents the concentration of the i-th factor in the current culture medium, and X ideal，j represents the concentration of the j-th factor in the optimized culture medium. The values of i and j are the same, and ∈ is a preset constant to prevent the denominator from being zero;
[0016] Obtain R i represents the ratio of the concentration of the i-th factor in the current culture medium to the concentration in the optimized culture medium;
[0017] If R i is equal to one, then the element a i is equal to one. If R i is greater than one, then the element a i is equal to K. If R i is less than one, then the element a i is equal to 1 / K, where K is a preset positive number;
[0018] Then calculate the influence coefficient w i of the i-th factor:
[0019] Next, introduce a non-linear adjustment factor:
[0020] F i = w i ·ΔX i ·(1 + α·ln(1 + ΔX i )); α is a preset adjustment factor, and F i represents the difference value of the concentration of the jth factor;
[0021] The calculation formula for the medium deviation index is:
[0022] γ represents a preset regulation factor, and CMDI represents the medium deviation index.
[0023] In a preferred embodiment, the acquisition logic of the plant adaptability index is as follows:
[0024] There are m plant growth characteristics in the plant applicability deviation characteristic data group, and the deviation of the rth plant growth characteristic is ΔG r , and the calculation formula is:
[0025] ∈ is a preset constant to prevent the denominator from being zero; G current，r represents the value corresponding to the rth plant growth characteristic in the current medium, and G target，r represents the value corresponding to the rth plant growth characteristic in the optimized medium;
[0026] Normalize each characteristic deviation ΔG r into a characteristic adaptability function to obtain the characteristic adaptability value H r : H r = 1 - tanh(|ΔG r |);
[0027] Calculate the interactive adaptability value I ry : I ry = δ ry ·(H r ·H y ); H r ·H y represents the interactive contribution of the adaptability functions corresponding to the plant growth characteristics r and y, and δ ry represents the preset interactive weight between the rth and yth plant growth characteristics;
[0028] Perform non-linear adjustment on each characteristic adaptability value to obtain the adjusted characteristic adaptability value:
[0029] β represents a preset adjustment coefficient; H' r represents the adjusted characteristic adaptability value;
[0030] The calculation formula of the plant adaptability index is as follows:
[0031] PAI represents the plant adaptability index.
[0032] In a preferred embodiment, the buffer determination refers to:
[0033] Taking the plant adaptability index and the culture medium deviation index as the input data of fuzzy inference, and taking the culture medium replacement type as the output data of fuzzy inference, it is obtained that the culture medium replacement type is high fitness optimization or low fitness optimization. When the culture medium replacement type is high fitness optimization, the duration of the buffer stage is zero;
[0034] When the culture medium replacement type is low fitness optimization, the duration of the buffer stage is:
[0035] T 0 represents the preset basic duration, f1 and f2 are both preset non-zero optimization coefficients, and T buffer represents the determined buffer stage duration.
[0036] In a preferred embodiment, the buffer stage refers to: before replacing the current culture medium with the optimized culture medium corresponding to parameter group two, using a preset buffer culture medium for buffer adjustment.
[0037] The preset buffer culture medium contains basic nutrients, plant hormones and mineral elements, and is used to maintain the normal growth of the fire blight-resistant pear rootstock tissue.
[0038] The technical effects and advantages of the present invention:
[0039] The present invention uses the plant adaptability index and the culture medium deviation index to scientifically judge the risk of culture medium switching. By dynamically setting the duration of the buffer stage and using a preset buffer culture medium, the sensitivity of plants to culture medium mutations is reduced, and the stress response caused by direct replacement is avoided. The buffer stage enables plants to complete the culture medium switching in an environment with good adaptability, reducing problems such as metabolic disorders and growth stagnation.
[0040] The present invention meets the specific requirements of plants at different growth stages through the optimized culture medium components, enhances the ability to resist fire blight, provides an environment suitable for the growth of pear rootstock tissue, ensures the healthy growth of plants and the complete development of roots, and improves the success rate of tissue culture. Using a classification model to judge whether the culture medium needs to be replaced, and predicting the optimized culture medium components through a regression model, the entire culture medium replacement process is intelligently managed. BRIEF DESCRIPTION OF THE DRAWINGS
[0041] For the convenience of those skilled in the art to understand, the present invention will be further described below with reference to the accompanying drawings;
[0042] Figure 1 This is the schematic diagram of the environmental control method for the tissue culture of fire blight-resistant pear rootstocks in the present invention. Detailed implementation manners
[0043] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without making creative efforts shall fall within the protection scope of the present invention.
[0044] Refer to Figure 1 The following embodiments are obtained:
[0045] Embodiment 1:
[0046] The environmental control method for the tissue culture of fire blight-resistant pear rootstocks includes the following steps: during the tissue culture of fire blight-resistant pear rootstocks, obtain the current culture medium usage data set caused by plant tissue metabolism and the plant tissue growth data set; the current culture medium usage data set reflects the dynamic changes of the components in the culture medium, such as nutrient consumption, pH value fluctuation, etc. The plant tissue growth data set can reflect the current growth status of the plant, such as growth rate, leaf size, root system expansion, etc., and can provide key data for subsequent analysis and model input, providing a scientific basis for optimizing the culture conditions.
[0047] Substitute the current culture medium usage data set and the plant tissue growth data set into the pre-trained machine learning model together, output the result of whether to replace the culture medium, and when the culture medium needs to be replaced, predict the components of the culture medium to obtain the parameter set two corresponding to the optimized culture medium; use the machine learning model instead of manual judgment to reduce human error and improve decision-making efficiency. The classification model can accurately identify whether the culture medium meets the current plant growth requirements by learning a large amount of historical data, avoiding unnecessary replacement or missing the best replacement time. Once replacement is needed, the regression model can quickly predict the parameter set of the optimized culture medium, providing direct guidance for subsequent operations.
[0048] Extract features and analyze the parameter set two corresponding to the optimized culture medium and the parameter set one corresponding to the current culture medium to obtain the analysis result; analyze the component differences between the current culture medium and the optimized culture medium to clarify the specific direction that needs to be adjusted. Extract the culture medium component difference feature data set and the plant applicability deviation feature data set, and calculate the culture medium deviation index and the plant adaptability index respectively for the subsequent buffering stage and replacement decision. Through the difference and adaptability analysis, clarify the possible bottlenecks in the process of adjusting the culture medium to ensure the scientific nature of replacing the culture medium.
[0049] Based on the analysis results, buffer determination is performed to determine the duration of the buffer phase to be experienced before replacing the current culture medium with the optimized culture medium corresponding to parameter set two and execute it. Plants may experience discomfort or growth disorders due to sudden changes in the culture medium. The buffer phase allows plants to gradually adapt to the new environment. Based on the plant adaptability index and the culture medium deviation index, the duration of the buffer phase is intelligently determined, which can not only avoid wasting time but also ensure the health of the plants. During the buffer phase, a preset buffer culture medium is used to reduce the adverse effects of culture medium switching on plant growth by maintaining the basic growth requirements of the plants.
[0050] The objectives of the present invention: Traditional culture methods mainly rely on manual experience to judge whether the culture medium needs to be replaced, which may lead to decision-making delays or misjudgments and affect the healthy growth of plants. Many culture systems replace the culture medium according to a fixed cycle, ignoring the real-time growth state of the plants and the specific changes in the culture medium, which easily causes resource waste or plant adaptability problems. Fire blight is a serious plant disease that poses a great threat to the healthy growth of pear tree rootstocks. Tissue culture is an important means of cultivating healthy rootstocks, but it is necessary to accurately control the culture medium during the culture process to improve disease resistance. Pear rootstocks may show significant growth fluctuations when the composition of the culture medium changes, and effective management is required through dynamic monitoring and a buffer mechanism. By introducing a machine learning model, real-time monitoring and dynamic adjustment of the culture medium state are achieved, greatly improving the efficiency and accuracy of tissue culture. By scientifically determining the timing of culture medium replacement, resource waste caused by premature or late replacement is avoided. By dynamically adjusting the buffer phase and optimizing the culture medium composition, the growth potential and disease resistance of plants are maximized.
[0051] The pre-trained machine learning model is a joint model, which is composed of a pre-trained classification model and a regression model. The classification model is used to judge whether the current culture medium needs to be replaced. When the classification model outputs the result that the current culture medium needs to be replaced, the prediction mechanism is triggered, the regression model is enabled, and the components of the culture medium to be replaced are predicted to obtain parameter set two corresponding to the optimized culture medium.
[0052] Through learning historical data, the classification model can identify the complex relationship between the culture medium state and the plant growth state. It analyzes the input current culture medium usage data and plant growth data to judge whether the culture medium can no longer meet the growth requirements of the plants, and then outputs a classification result of "needs to be replaced" or "does not need to be replaced". Logistic regression, support vector machine (SVM), random forest, or a binary classification model in deep learning (such as a neural network with binary cross-entropy loss) can be used. Through machine learning training, the classification model can accurately identify the timing of culture medium replacement, avoiding plant growth disorders caused by delayed replacement. Automated decision-making reduces the need for human intervention and shortens the judgment cycle. The self-learning ability of the model can continuously improve the accuracy of judgment through data accumulation.
[0053] When the classification model determines that the culture medium needs to be replaced, the regression model is triggered to predict the optimal ratio of the culture medium components. The model selection can adopt random forest regression, gradient boosting regression (such as XGBoost or LightGBM), or a neural network regression model. The regression model calculates the culture medium parameters most suitable for the current plant growth by analyzing the growth performance of the plant under different culture medium conditions in historical data. The regression model can predict the optimized culture medium components based on the actual growth state of the plant and the usage of the current culture medium, avoiding relying on fixed or empirical formulas. The growth requirements and culture medium usage of different plants may vary, and the regression model can adjust the optimization strategy according to real-time data. By providing a more suitable combination of culture medium components for the plant, it promotes its faster and healthier growth.
[0054] The combined model combines the functions of the classification model and the regression model. It can not only judge whether the culture medium needs to be replaced but also provide an optimization plan when replacement is needed. The classification model is responsible for making the decision on whether to replace, and the regression model is responsible for designing the optimized culture medium components. The two combined constitute a complete culture medium management system. The combined model in the existing technology effectively solves the problems in traditional culture medium management, significantly improving the culture efficiency, accuracy, and plant growth performance. By using the classification model to judge whether the culture medium needs to be replaced, unnecessary resource waste is avoided; by using the regression model to design the optimized culture medium components, the adaptability and growth rate of the plant are improved, providing an important technical support for the intelligent culture environment control.
[0055] The feature extraction and analysis of the parameter group two corresponding to the optimized culture medium and the parameter group one corresponding to the current culture medium refer to: based on the parameter group two corresponding to the optimized culture medium and the parameter group one corresponding to the current culture medium, respectively extract the culture medium component difference feature data group and the plant applicability deviation feature data group, and then perform difference analysis on the culture medium component difference feature data group to generate a culture medium deviation index, and perform deviation analysis on the plant applicability deviation feature data group to generate a plant adaptability index. It should be noted that: the parameter group one contains the specific component data in the current culture medium and the plant growth performance data in the current culture medium, and the parameter group two contains the specific component data in the optimized culture medium and the plant growth performance data in the optimized culture medium. The plant growth performance data in the optimized culture medium can be extracted from the information pre-stored in the database or obtained through simulation.
[0056] Extract the differential characteristic data set of the culture medium components, which can clearly quantify the specific component differences between the current culture medium and the target optimized culture medium, such as the differences in nutrient concentrations and mineral element contents. The culture medium deviation index provides a global numerical indicator by comprehensively analyzing the deviation degrees of all components, which is used to quickly evaluate the overall difference between the current culture medium and the target culture medium. The culture medium deviation index can be used for the determination in the subsequent buffer stage. The greater the difference, the greater the necessity and longer the duration of the buffer stage.
[0057] The Plant Adaptability Index (PAI) is mainly used to quantify the growth performance differences of plants under the conditions of the current culture medium (parameter group one) and the optimized culture medium (parameter group two). The larger the PAI value, the greater the difference in the growth performance of the plant in the optimized culture medium compared to the current culture medium; while the smaller the PAI value, the closer the growth performances of the two are. A larger PAI value indicates a significant difference in the growth performance of the plant under the conditions of the optimized culture medium and the current culture medium. The difference may be due to significant changes in some components in the optimized culture medium, and the plant needs time to gradually adapt. Otherwise, direct replacement may lead to growth stress, resulting in growth stagnation or damage to health. A smaller PAI value indicates that the optimized culture medium has a similar growth performance to the current culture medium for the plant, and direct replacement of the culture medium has less impact on the adaptability of the plant, and even the buffer stage can be omitted. The PAI value can provide a quantitative basis for whether a buffer stage is needed. The greater the difference, the longer the duration of the buffer stage needs to be increased. A larger PAI value indicates a higher risk of direct replacement and requires adjustment to mitigate the stress response.
[0058] The acquisition logic of the culture medium deviation index is as follows:
[0059] There are n culture medium factors in the differential characteristic data set of the culture medium components, and the deviation of the i-th factor is ΔX i , and the calculation formula is:
[0060] X act，i represents the concentration of the i-th factor in the current culture medium, and X ideal，j represents the concentration of the j-th factor in the optimized culture medium. The values of i and j are the same, and ∈ is a preset constant to prevent the denominator from being zero; calculate the relative deviation between the actual value and the optimized target value of the i-th component (i.e., factor) in the current culture medium, avoiding the calculation imbalance caused by simply relying on the absolute difference. Ensure that both positive and negative deviations are included in the calculation through the absolute value to comprehensively reflect the actual differences of the factors.
[0061] Obtain R i represents the ratio of the concentration of the i-th factor in the current culture medium to the concentration in the optimized culture medium;
[0062] If R i is equal to one, then the element a is obtainedi equals one if R i is greater than one, then element a is obtained i equals K if R i is less than one, then element a is obtained i equals 1 / K, where K is a preset positive number; by converting the ratio to 1, K, or 1 / K, corresponding to the ideal state, excess, or deficiency state respectively, it is convenient for subsequent processing. By introducing K, the sensitivity to too large or too small deviations is increased, making it more instructive in deviation analysis.
[0063] Then calculate the influence coefficient w of the i-th factor i : Ensure that the weight of each factor is within the range of 0 - 1, and the sum of the weights of all factors is 1, which is convenient for fairly evaluating the contribution of different factors to the overall deviation. According to the importance of the factor, i.e., the value of element a i to assign weights, more scientifically highlighting the influence of key components and providing a flexible way of weight assignment.
[0064] Then introduce a non-linear adjustment factor:
[0065] F i = w i ·ΔX i ·(1 + α·ln(1 + ΔX i )); α is a preset adjustment factor, F i represents the difference value of the concentration of the j-th factor; ΔX i provides the deviation basis for each factor. By w i highlighting the contribution ratio of different factors in the overall deviation, introducing the non-linear term α·ln(1 + ΔX i ), the factors with larger deviations are significantly amplified, increasing the sensitivity to abnormal factors.
[0066] The calculation formula for the medium deviation index is:
[0067] γ represents a preset regulation factor, and CMDI represents the medium deviation index. Combining the deviations of all single factors into a total index can intuitively reflect the overall deviation of the medium, weakening the contribution of factors with smaller deviations, making the overall index more sensitive to factors with large deviations. The adjustment parameter γ can adjust the intensity of index attenuation as needed to balance the influence of single factors and overall deviations. The larger the medium deviation index, the greater the impact that the medium replacement may cause to the plant, and it is necessary to gradually transition through a buffer stage, which means that the environmental difference between the current medium and the optimized medium is obvious. If directly switched to the optimized medium, the plant may show a decline in adaptability or growth inhibition.
[0068] The acquisition logic of the plant adaptability index is as follows:
[0069] There are m plant growth characteristics in the plant applicability deviation characteristic data group, and the deviation of the r-th plant growth characteristic is ΔG r , and the calculation formula is:
[0070] ∈ is a preset constant to prevent the denominator from being zero; G current，r represents the value corresponding to the r-th plant growth characteristic in the current culture medium, and G target，r represents the value corresponding to the r-th plant growth characteristic in the optimized culture medium; by calculating the deviation of plant growth characteristics (such as root length, leaf area, growth rate, etc.) in the current culture medium and the optimized culture medium, the difference before and after optimization is directly reflected.
[0071] Normalize each characteristic deviation ΔG r into a characteristic adaptability function to obtain the characteristic adaptability value H r : H r = 1 - tanh(|ΔG r |); use the hyperbolic tangent function tanh to map the deviation ΔG r to the adaptability value range of [0,1]. The non-linear characteristic of the tanh function can weaken the extreme influence of large deviations on the result and highlight the role of medium and small deviations at the same time. H r being 1 means almost no deviation, and the plant has the best adaptability to this characteristic. The closer H r is to 0, the greater the deviation, and the greater the difference in the plant's adaptability to this characteristic.
[0072] Calculate the interactive adaptability value I ry : I ry = δ ry ·(H r ·H y ); H r ·H y represents the interactive contribution of the adaptability functions corresponding to the plant growth characteristics r and y. δ ry represents the preset interaction weight between the r-th and y-th plant growth characteristics; there may be an interaction between characteristics (such as the relationship between photosynthesis and leaf area), and this formula quantitatively evaluates these synergies or conflicts. H r ·H y represents the product of the independent adaptability values of two characteristics, reflecting their interactive contribution. δ ry adjusts the importance of the interactive effects between different characteristics and flexibly adapts to the characteristic requirements of different plants.
[0073] Perform non-linear adjustment on each characteristic adaptability value to obtain the adjusted characteristic adaptability value:
[0074] β represents a preset adjustment coefficient; H′ r represents the adjusted characteristic adaptability value; the non - linear adjustment introduces the adjustment coefficient β to dynamically adjust the adaptability value, so that characteristics with larger deviations receive more attention.
[0075] The calculation formula of the plant adaptability index is:
[0076] PAI represents the plant adaptability index. The first part is the single - characteristic contribution, H′ r ·ln(1 + |ΔG r |) comprehensively considers the characteristic adaptability value and the deviation value. ln(1 + |ΔG r |) amplifies the contribution of characteristics with larger deviations. Normalize all characteristics to ensure that the result is not affected by the number of characteristics. The second part is the interaction - effect contribution. Calculate the interaction adaptability values between all characteristics to reflect the overall synergy effect. Combine the single - characteristic and interaction - effect to comprehensively evaluate the difference in the growth performance of plants in the culture media before and after optimization. A larger PAI value indicates that there are significant differences in the growth performance of plants in the optimized culture medium and the current culture medium, which may be due to significant changes in some key components (such as nutrients, mineral elements, hormone concentrations, etc.) in the optimized culture medium. Components with large differences may have important effects on the growth characteristics of plants (such as root length, leaf area, growth rate, etc.), resulting in significant differences in the growth performance of plants compared with the current culture medium. When PAI is large, it is necessary to design a buffer stage to reduce the stress response of plants due to environmental mutations. The duration of the buffer stage should be dynamically optimized according to the size of PAI.
[0077] The buffer determination refers to:
[0078] Take the plant adaptability index and the culture - medium deviation index as the input data of fuzzy inference, and take the culture - medium replacement type as the output data of fuzzy inference to obtain that the culture - medium replacement type is high - fitness optimization or low - fitness optimization. When the culture - medium replacement type is high - fitness optimization, the duration of the buffer stage is zero;
[0079] When the culture - medium replacement type is low - fitness optimization, the duration of the buffer stage is:
[0080] T 0 represents a preset basic duration, f1 and f2 are both preset non - zero optimization coefficients, and T buffer represents the determined duration of the buffer stage.
[0081] Convert the input data (such as plant adaptability index and medium deviation index) into membership values of fuzzy sets. The range of membership values is usually between 0 and 1, indicating the degree to which the data belongs to a certain fuzzy set. For example: An adaptability index of 0.8 may correspond to a membership degree of 0.7 for "high adaptability" and a membership degree of 0.3 for "medium adaptability". Role of trigonometric functions: The membership functions of fuzzy sets usually adopt forms such as triangular membership functions or trapezoidal membership functions, and the input values are fuzzified through these functions. The mathematical form of the triangular membership function is simple and the calculation is efficient, and it is often used in fuzzy inference systems.
[0082] Use the preset fuzzy logic rules to match the fuzzified input values with the rule base. For example: Rule: "If the adaptability index is low and the deviation index is high, then the medium replacement type is low fitness optimization."
[0083] Calculate the contribution of the input fuzzy set to the output fuzzy set according to the rules. Common methods: Max - Min method: Use the "take the minimum" operation on the membership value of the input and the membership degree of the conditional part in the rule. Fuzzy operators: Use logical operators (such as intersection, union) to combine the input fuzzy sets. Convert the fuzzy inference result into a specific output value (such as the duration of the buffer stage). Common defuzzification methods: Centroid method: Calculate the centroid of the fuzzy set as the final output value. Maximum membership degree method: Select the fuzzy set with the maximum membership degree as the output.
[0084] Plant adaptability index: Reflects the growth performance difference of plants under the current medium and the optimized medium. Medium deviation index: Reflects the difference degree in composition between the current medium and the optimized medium. According to the adaptability index and the deviation index, output the medium replacement type: High fitness optimization: Direct switch, no buffer stage required. Low fitness optimization: A buffer stage needs to be set.
[0085] The regulation of the buffer stage duration aims to ensure that plants can smoothly transition to the optimized culture medium through scientific timing arrangements and culture medium adjustments. In cases where plants have poor adaptability to the optimized culture medium, direct switching may lead to reduced plant growth rates, abnormal growth, or even growth failure. By setting up the buffer stage, the impact of environmental mutations on plants can be reduced. The length of the buffer duration is dynamically adjusted according to the plant adaptability index and the culture medium deviation index: when the difference is small, the buffer stage duration is short or no buffer is required. When the difference is large, the buffer stage duration is correspondingly extended to provide sufficient adaptation time for the plants. Through reasonable buffer stage design, plant growth failure caused by insufficient buffering is avoided, and time and resource waste caused by overly long buffering are also avoided. Dynamically regulating the duration can balance plant adaptability and culture efficiency to the greatest extent. The buffer stage not only reduces the risk of growth stress but also provides time and space for plants to gradually adapt to the optimized culture medium, ensuring the healthy growth of plants in the optimized culture medium. The regulation of the buffer stage duration relies on two scientific indicators, the plant adaptability index and the culture medium deviation index, and combines fuzzy reasoning to achieve intelligent and automated management of culture medium replacement, and can optimize the buffer strategy in a timely manner according to the specific growth performance of plants.
[0086] The buffer stage refers to: before replacing the current culture medium with the optimized culture medium corresponding to parameter group two, buffer adjustment is carried out using a preset buffer culture medium. The preset buffer culture medium contains basic nutrients, plant hormones, and mineral elements, and is used to maintain the normal growth of the fire blight-resistant pear rootstock tissue. Plants are extremely sensitive to changes in the culture medium. Direct switching to the optimized culture medium may lead to mutations in the nutrient environment and chemical balance, thereby triggering growth stress (such as reduced growth rate, metabolic imbalance, cell damage, etc.). Through the buffer stage, by using a preset buffer culture medium with balanced and suitable components, plants can gradually adapt to the changes in the culture medium and reduce stress responses. While improving the success rate of culture medium replacement, it ensures the growth stability and disease resistance of plants, especially for the culture of fire blight-resistant pear rootstock tissue.
[0087] The preset buffer culture medium contains basic nutrients required for plant growth (such as nitrogen, phosphorus, potassium, etc.), plant hormones (such as auxin, cytokinin), and mineral elements (such as calcium, magnesium, iron, etc.). During the buffer stage, the buffer culture medium can provide necessary nutrients and regulatory factors for plants, maintain normal metabolism and growth, ensure that plants can obtain sufficient nutrients and regulatory factors during the buffer stage, maintain basic physiological functions, and create good preconditions for the switch to the optimized culture medium. The drastic changes during the culture medium switch may lead to growth disorders or even failure of plants, especially for sensitive plants resistant to fire blight. The buffer stage reduces the risk of this process through gradual adaptation, significantly improving the stability and success rate of culture. It provides higher controllability for the culture of pear rootstock tissue, especially in the important application scenario of fire blight resistance.
[0088] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data and performing software simulations to get a formula that is closest to the actual situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.
[0089] It should be understood that in various embodiments of the present application, the magnitudes of the sequence numbers of the above processes do not mean the order of execution. The order of execution of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present application.
[0090] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present application.
[0091] Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments, and will not be elaborated herein.
[0092] The above is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present application, and all should be covered by the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.
Claims
1. A method for controlling the environment of pear rootstock tissue culture for resistance to fire blight of pear, characterized in that: The following steps are involved: In the tissue culture process of pear rootstock resistant to fire blight, a data set of current medium usage caused by plant tissue metabolism and a data set of plant tissue growth are obtained; Substitute the current culture medium usage data group and the plant tissue growth data group into the pre-trained machine learning model, output the result of whether to replace the culture medium, and when the culture medium needs to be replaced, predict the culture medium composition to obtain the second parameter group corresponding to the optimized culture medium; Extract and analyze the characteristics of parameter group 2 corresponding to the optimized culture medium and parameter group 1 corresponding to the current culture medium to obtain an analysis result; A buffer determination is performed based on the analysis results, and the duration of the buffer phase to be experienced before the current culture medium is replaced with the optimized culture medium corresponding to parameter group two is determined and executed.
2. the fire blight-resistant pear rootstock tissue culture environment control method according to claim 1, is characterized in that, The pre-trained machine learning model is a joint model, which consists of a pre-trained classification model and a regression model.
3. the fire blight-resistant pear rootstock tissue culture environment control method according to claim 2, is characterized in that, The classification model is used to determine whether the current culture medium needs to be replaced. When the classification model outputs the result that the current culture medium needs to be replaced, the prediction mechanism is triggered, the regression model is enabled, and the components of the culture medium that need to be replaced are predicted to obtain parameter group 2 corresponding to the optimized culture medium.
4. the fire blight resistant pear rootstock tissue culture environment control method according to claim 3, is characterized in that, Extracting and analyzing the characteristics of parameter group 2 corresponding to the optimized culture medium and parameter group 1 corresponding to the current culture medium means: Based on parameter group 2 corresponding to the optimized culture medium and parameter group 1 corresponding to the current culture medium, the culture medium component difference characteristic data group and the plant suitability deviation characteristic data group are extracted respectively, and then a difference analysis is performed on the culture medium component difference characteristic data group to generate a culture medium deviation index, and a deviation analysis is performed on the plant suitability deviation characteristic data group to generate a plant adaptability index.
5. The method for controlling the environment of the fire blight resistant pear rootstock tissue culture according to claim 4, wherein: The logic for obtaining the culture medium deviation index is: There are n culture medium factors in the culture medium component difference characteristic data set, and the deviation of the i-th factor is ΔX i , the calculation formula is: X act，i表 indicates the concentration of the i-th factor in the current culture medium, X ideal，j represents the concentration of the jth factor in the optimized culture medium, i and j have the same value, and ∈ is a preset constant to prevent the denominator from being zero; Get R i It represents the ratio of the concentration of the i-th factor in the current culture medium to the concentration in the optimized culture medium; If R i is equal to one, then we get element a i is equal to one, if R i If it is greater than one, we get element a i Equal to K, if R i If it is less than one, we get element a i Equal to one-K, where K is a preset positive number; Then calculate the influence coefficient w of the i-th factor i : Then introduce the nonlinear adjustment factor: F i =w i ΔX i ·(1+α·ln(1+ΔX i )); α is the preset adjustment factor, F i represents the difference value of the concentration of the jth factor; The calculation formula of medium deviation index is: γ represents the preset regulatory factor, and CMDI represents the culture medium deviation index.
6. The method for controlling the environment of the fire blight resistant pear rootstock tissue culture according to claim 5, wherein: The logic for obtaining the plant adaptability index is: There are m plant growth characteristics in the plant suitability deviation feature data set, and the deviation of the rth plant growth characteristic is ΔG r , the calculation formula is: ∈ is a preset constant to prevent the denominator from being zero; G current，r represents the value of the rth plant growth characteristic in the current culture medium, G target，r represents the value corresponding to the rth plant growth characteristic in the optimized culture medium; Each characteristic deviates from ΔG r Normalized to the characteristic adaptability function, the characteristic adaptability value H is obtained r :H r =1-tanh(|ΔG r |); Calculate the interactive fitness value I between features ry :I ry =δ ry ·(H r ·H y );H r ·H y represents the interactive contribution of plant growth characteristics r and y to the fitness function, δ ry Represents the preset interaction weight between the rth and yth plant growth characteristics: Perform nonlinear adjustment on each characteristic adaptability value to obtain the adjusted characteristic adaptability value: β represents the preset adjustment coefficient; H′ r Indicates the adjusted characteristic adaptability value: The calculation formula of plant adaptability index is: PAI stands for Plant Adaptability Index.
7. The method for controlling the environment of the fire blight resistant pear rootstock tissue culture according to claim 6, wherein: Buffer determination refers to: The plant adaptability index and the culture medium deviation index are used as the input data of fuzzy reasoning, and the culture medium replacement type is used as the output data of fuzzy reasoning, and the culture medium replacement type is obtained as high fitness optimization or low fitness optimization. When the culture medium replacement type is high fitness optimization, the buffer stage duration is zero; When the medium replacement type is low fitness optimization, the buffer phase duration is: T0 represents the preset basic duration, f1 and f2 are both preset non-zero optimization coefficients, T buffer Indicates the determined buffering phase duration.
8. The method for controlling the environment of the fire blight resistant pear rootstock tissue culture according to claim 7, wherein: The buffering stage refers to: before replacing the current culture medium with the optimized culture medium corresponding to parameter group two, buffering adjustment is performed using the preset buffering culture medium.
9. The method for controlling the environment of the fire blight resistant pear rootstock tissue culture according to claim 8, wherein: The preset buffered culture medium contains basic nutrients, plant hormones and mineral elements to maintain the normal growth of pear rootstock tissues resistant to fire blight.