A method for assessing the dynamic stress resistance of oilseed crops based on growth stage division

Abstract

This application provides a method for assessing the dynamic stress resistance of oilseed crops based on growth stage division, belonging to the field of agricultural information technology. The method includes: dividing growth stages according to physiological cycles and collecting field, environmental, and disaster early warning data in stages; constructing an evaluation index system for the dominant stress resistance traits at each stage; calculating the dynamic weights of the indicators using a time-series entropy weighting method that integrates stage importance coefficients, thereby obtaining a stage stress resistance score; constructing a state transition model based on the stage scores to characterize the stress resistance evolution path and output a state transition stationarity score; and generating a dynamic stress resistance index throughout the entire growth period by combining the stage scores and the stationarity score, achieving a dynamic and quantitative evaluation of stress resistance throughout the entire growth cycle.

Description

Technical Field

[0001] This application belongs to the field of agricultural information technology, specifically relating to a method for assessing the dynamic stress resistance of oilseed crops based on growth stage division. Background Technology

[0002] In existing crop stress resistance evaluation systems, static or uniform evaluation models covering the entire growth period are typically used. These models integrate variety approval data, field measurement data, and meteorological disaster information to construct a comprehensive stress resistance index to assess the variety's stress resistance. While such methods can reflect the overall stress resistance level of a variety to some extent, they still have significant limitations: most evaluation models treat the entire growth period of a crop as a homogeneous whole, ignoring the inherent laws that may lead to significant changes in stress resistance traits at different growth stages (such as seedling stage, flowering stage, pod-setting stage, and maturity stage). For example, the same variety may be sensitive to low temperatures during the seedling stage, but may exhibit strong lodging resistance at maturity.

[0003] In existing technologies, stress resistance evaluation often relies on data from a specific period or critical stage, lacking a continuous characterization of the dynamic evolution of stress resistance traits throughout the entire crop growth cycle. This makes it difficult for evaluation results to accurately reflect the true adaptability of varieties to stage-specific disasters in actual production, and also fails to provide effective guidance for refined cultivation management or stress resistance breeding at different growth stages. Summary of the Invention

[0004] This application provides a method for assessing the dynamic stress resistance of oilseed crops based on growth stage division, in order to solve one of the aforementioned technical problems.

[0005] The technical solution adopted in this application is as follows:

[0006] This application provides a method for assessing the dynamic stress resistance of oilseed crops based on growth stage division, including:

[0007] Acquire field measured data, environmental monitoring data and disaster early warning information of the target crop at each preset growth stage throughout its entire growth period, wherein each preset growth stage is pre-defined according to the crop's physiological cycle;

[0008] Based on the set of dominant stress resistance traits corresponding to each growth stage, a stress resistance evaluation index system for each stage is constructed. For each growth stage, the time entropy weight method that integrates the stage importance coefficient is used to calculate the dynamic weight of each evaluation index in that stage.

[0009] Based on the indicator data and their dynamic weights within each stage, the stage resilience score for each growth stage is calculated. Based on the stage resilience scores of all growth stages, a resilience state transition model is constructed to describe the evolution path of the resilience state between each stage and output the state transition stability score.

[0010] By combining the stage-specific stress resistance scores at each growth stage with the state transition stability scores, a dynamic stress resistance index for the target crop throughout its growth process is generated.

[0011] According to one embodiment of this application, the dynamic weight of each evaluation index within a given stage is calculated using a time-series entropy weighting method that incorporates stage importance coefficients. Specifically, the calculation is performed using the following formula:

[0012]

[0013] in, For the first The first growth stage The dynamic weights of the evaluation indicators The entropy weight is calculated based on indicator data. The first [stage] determined based on the historical frequency of disaster occurrence and the duration of each stage The importance coefficient of each growth stage. and For the first Harmonization coefficients for each growth stage.

[0014] According to one embodiment of this application, the stress resistance state transition model is a hidden Markov model or a state-space model; the evolutionary path is used to identify the stress-vulnerable stage of the target crop during its growth period.

[0015] According to one embodiment of this application, the formula for calculating the dynamic stress resistance index throughout the entire process is as follows:

[0016]

[0017] in, The dynamic resilience index throughout the entire process, This represents the total number of growth stages. For the first The overall weight of each growth stage For the first Stage-specific stress resistance scores for each growth stage Score the state transition stationarity. These are the preset transfer weight coefficients.

[0018] According to one embodiment of this application, it also includes:

[0019] Based on the dynamic resilience index and the evolution path output by the resilience state transition model, a visual chart containing resilience evolution curves and / or stage comparison radar charts is generated.

[0020] A second aspect of this application provides a dynamic stress resistance assessment system for oilseed crops based on growth stage division, comprising:

[0021] The data acquisition module is used to acquire field measured data, environmental monitoring data and disaster early warning information of the target crop at each preset growth stage throughout its entire growth period. The preset growth stages are predefined according to the crop's physiological cycle.

[0022] The indicator system construction module is used to construct an indicator system for evaluating the stress resistance of each growth stage based on the set of dominant stress resistance traits corresponding to each growth stage.

[0023] The weight calculation module is used to calculate the dynamic weight of each evaluation index within each growth stage using the time-series entropy weight method that incorporates the stage importance coefficient.

[0024] The stage scoring module is used to calculate the stage stress resistance score for each growth stage based on the indicator data and their dynamic weights within each stage.

[0025] The state transition analysis module is used to construct a stress resistance state transition model based on the stage stress resistance scores of all growth stages, to describe the evolution path of the stress resistance state between each stage, and to output the state transition stability score.

[0026] The index integration module is used to combine the stage stress resistance scores of each growth stage and the state transition stability scores to generate the full-process dynamic stress resistance index of the target crop.

[0027] According to one embodiment of this application, the weight calculation module is specifically used to calculate dynamic weights using the following formula:

[0028]

[0029] in, For the first The first growth stage The dynamic weights of the evaluation indicators The entropy weight is calculated based on indicator data. The first [stage] determined based on the historical frequency of disaster occurrence and the duration of each stage The importance coefficient of each growth stage. and For the first Harmonization coefficients for each growth stage.

[0030] According to one embodiment of this application, the model constructed by the state transition analysis module is a hidden Markov model or a state-space model; the evolution path is used to identify the stress-resistance and vulnerability stages of the target crop during its growth period.

[0031] According to one embodiment of this application, the calculation formula for the full-process dynamic resilience index generated by the index integration module is as follows:

[0032]

[0033] in, The dynamic resilience index throughout the entire process, This represents the total number of growth stages. For the first The overall weight of each growth stage For the first Stage-specific stress resistance scores for each growth stage Score the state transition stationarity. These are the preset transfer weight coefficients.

[0034] According to one embodiment of this application, it also includes:

[0035] The visualization module is used to generate visualization charts containing resilience evolution curves and / or stage comparison radar charts based on the evolution path output by the full-process dynamic resilience index and the resilience state transition model.

[0036] Due to the adoption of the above technical solution, the beneficial effects achieved by this application are as follows:

[0037] This application, by pre-dividing crop growth stages and constructing an evaluation system in stages, can accurately capture the specific performance of stress resistance traits at each stage, thereby achieving a continuous and dynamic characterization of the evolution of variety stress resistance over time, making the evaluation results more consistent with the actual growth patterns of the crop.

[0038] By constructing an evaluation index system corresponding to the dominant stress resistance traits for each growth stage and introducing a stage importance coefficient based on historical disaster data, the evaluation system can focus on reflecting the variety's resistance to major stress factors at different stages, which helps to identify its stress-vulnerable stages.

[0039] The time-series entropy weighting method, which incorporates the importance coefficient of each stage, is used to calculate the dynamic weight of the indicators within each stage. This method not only considers the objective differences in the indicator data but also incorporates the relative importance of each growth stage in the overall stress resistance evaluation, making the weight allocation more reasonable and in line with agronomic practice.

[0040] By constructing a stress resistance state transition model, analyzing the transition relationship and evolution path between stress resistance scores at each stage, and outputting a state transition stability score, we can comprehensively evaluate the coordination and stability of the variety's stress resistance from the perspective of temporal correlation, providing a deeper basis for judging its adaptability throughout the entire growth period.

[0041] By integrating the stress resistance scores and state transition stability scores at each stage, a dynamic stress resistance index (DARI) is generated. This index not only reflects the stress resistance performance of a variety at each stage, but also reflects the evolutionary quality of its stress resistance traits throughout the entire growth period, thus providing a more comprehensive and systematic quantitative evaluation result to support refined decision-making in variety breeding and cultivation management. Attached Figure Description

[0042] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0043] Figure 1 This is a flowchart illustrating a method for assessing the dynamic stress resistance of oil crops based on growth stage division, as provided in an embodiment of this application. Detailed Implementation

[0044] To more clearly illustrate the overall concept of this application, a detailed explanation is provided below with reference to the accompanying drawings.

[0045] Many specific details are set forth in the following description to provide a thorough understanding of this application. However, this application may also be implemented in other ways different from those described herein. Therefore, the scope of protection of this application is not limited to the specific embodiments disclosed below. It should be noted that, unless otherwise specified, the embodiments of this application and the features thereof can be combined with each other.

[0046] In this application, unless otherwise expressly specified and limited, the "above" or "below" of the second feature can mean that the first and second features are in direct contact, or that the first and second features are in indirect contact through an intermediate medium. In the description of this specification, references to terms such as "an embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described can be combined in any suitable manner in one or more embodiments or examples.

[0047] Example 1

[0048] like Figure 1 As shown, a method for assessing the dynamic stress resistance of oilseed crops based on growth stage division includes:

[0049] The target crop is obtained from field measured data, environmental monitoring data and disaster early warning information at each preset growth stage throughout its entire growth period. The preset growth stages are determined in advance according to the crop's physiological cycle.

[0050] As mentioned above, this step forms the data foundation and establishes the phased framework of this method. Its core lies in dividing the growth stages according to the crop's physiological cycle and collecting multi-source data in stages. Firstly, "pre-determining the growth stages according to the crop's physiological cycle" refers to dividing the target crop (such as rapeseed and soybean) into key growth periods that are generally recognized as having significantly different physiological characteristics and stress response mechanisms. Examples include seedling stage, budding stage, flowering stage, pod-setting stage, and maturity stage. This division is not a simple equal division of time, but a scientific periodization based on agronomic knowledge, aiming to ensure a relatively consistent focus for stress resistance evaluation within each stage. Secondly, "acquiring...field measured data, environmental monitoring data, and disaster early warning information" refers to the simultaneous collection of three types of data within each pre-defined growth stage window: 1) Field measured data: This refers to crop trait data (such as plant height, stem diameter, disease index, lodging area) and final yield and loss rate directly measured in experimental or production fields by manual methods or sensors; 2) Environmental monitoring data: This refers to environmental parameters (such as temperature, humidity, precipitation, wind speed, and soil moisture) continuously recorded through meteorological stations and sensor networks; 3) Disaster early warning information: This refers to forecasts or real-time alerts from meteorological or agricultural departments (such as frost warnings, rainstorm warnings, and pest and disease occurrence forecasts). These three types of data need to be aligned with the growth stage in both time and space, collectively forming the input basis for the stress resistance evaluation at that stage.

[0051] For example, let's take winter rapeseed as an example. First, define the growth stages: based on its physiological cycle, pre-divide it into five stages: S1 seedling stage (emergence - pre-wintering), S2 overwintering stage (beginning of overwintering - greening), S3 budding stage (greening - initial flowering), S4 flowering stage (initial flowering - final flowering), and S5 maturity stage (final flowering - harvest). Then, within the specific S3 budding stage, acquire data from multiple sources:

[0052] Field measurement data: During this stage, the incidence of sclerotinia stem rot, the sampling value of stem strength, and the plant density of rapeseed in the experimental field were measured and recorded regularly.

[0053] Environmental monitoring data: Through small weather stations deployed in the fields, daily minimum temperature, precipitation and relative humidity data are continuously acquired during this period.

[0054] Disaster early warning information: Access the regional agricultural meteorological service system to obtain "late spring cold" frost warning information and sclerotinia disease meteorological risk forecast issued during the budding stage.

[0055] All data are labeled with their corresponding stage identifier (S3) and specific date to form a staged dataset, which provides data support for the subsequent construction of a stress resistance evaluation index system for the S3 stage (such as focusing on cold resistance and disease resistance).

[0056] Based on the set of dominant stress resistance traits corresponding to each growth stage, a stress resistance evaluation index system for each stage is constructed. For each growth stage, the time entropy weight method, which integrates the stage importance coefficient, is used to calculate the dynamic weight of each evaluation index within that stage.

[0057] As mentioned above, this step is the core modeling step of this method. Its purpose is to establish a scientific evaluation framework for each growth stage that is both stage-specific and reflects the relative importance of indicators. First, "constructing a stress resistance evaluation index system for each growth stage based on the set of dominant stress resistance traits corresponding to each growth stage" means: based on each growth stage defined in the first step, combined with the main stress types that the crop is prone to encounter at that stage (e.g., seedling stage is prone to low temperature and drought stress, maturity stage is prone to lodging and disease), determining the stress resistance traits that need to be evaluated at that stage (i.e., the "set of dominant stress resistance traits," such as cold resistance, drought resistance, lodging resistance, etc.). Then, for each abstract trait, one or more specific quantitative indicators that can be directly observed or calculated are selected or designed, thus forming the index system for that stage. For example, the trait "lodging resistance" is concretized into measurable indicators such as "measured stem strength," "plant height," and "root development index." Secondly, "for each growth stage, the dynamic weights of each evaluation indicator within that stage are calculated using a time-series entropy weight method that incorporates stage importance coefficients" means that after the indicator system is established, a weight is assigned to each indicator to reflect its importance in the evaluation of that stage. This uses an improved entropy weight method, which is improved in two ways: first, the "time-series" characteristic, meaning that when calculating entropy weights, not only current period data is considered, but also the dispersion of historical data from the same period is taken into account, making the weights more reflective of the long-term distinguishing ability of the indicators; second, "incorporating stage importance coefficients," meaning that the final dynamic weights of the indicators are not entirely data-driven, but are harmonized by introducing a "stage importance coefficient" based on agronomic a priori knowledge. This coefficient is determined comprehensively based on the criticality of that growth stage to the final yield or quality formation throughout the entire growth period, as well as the frequency and severity of historical disasters at that stage. By combining the objectivity of data (entropy weights) with the subjectivity of expert knowledge (stage importance), dynamic weights that are more closely aligned with the realities of agricultural production are calculated.

[0058] For example, continuing the previous example of winter rapeseed, during the S2 overwintering period, the dominant stress is low-temperature freezing damage; therefore, its "dominant set of stress resistance traits" is determined to be cold resistance. Based on this trait, an evaluation index system for the S2 stage is constructed, which may include: field-measured indicators such as "change rate of leaf conductivity after low-temperature stress" (reflecting the degree of cell membrane damage) and "overwintering survival rate"; and environmental indicators such as "extreme minimum temperature within the stage" and "number of consecutive days ≤0℃". Subsequently, dynamic weights are calculated for these indicators. First, data on each indicator during the S2 overwintering period of the current year and several past years are collected, and the dispersion of each indicator data across different years is calculated (i.e., traditional entropy weighting). Second, experts or historical disaster data are used to assess the extreme importance of the "overwintering period" for the safe overwintering and final yield of winter rapeseed, assigning it a high "stage importance coefficient" (e.g., 0.7, while other non-critical stages may be 0.3). Finally, the entropy weight of each indicator is fused with the importance coefficient of that stage using a preset algorithm (such as linear weighting) to generate the final dynamic weight of each indicator in stage S2. For example, the "overwintering survival rate" will have a significantly higher dynamic weight than other indicators because it directly reflects the final result and has large fluctuations in data over the years (high entropy weight) and high stage importance.

[0059] Based on the indicator data and their dynamic weights within each stage, the stage stress resistance score for each growth stage is calculated.

[0060] Based on the stage resilience scores of all growth stages, a resilience state transition model is constructed to describe the evolution path of resilience states between stages and output the state transition stability score.

[0061] As mentioned above, this step is a crucial link in moving from phased static assessment to dynamic correlation analysis throughout the entire growth period, encompassing two levels: "intra-phase synthesis" and "inter-phase correlation." First, "calculating the stage-specific stress resistance score for each growth stage" means that for each defined growth stage with a constructed indicator system and dynamic weights, the actual observed values ​​(or normalized values) of each indicator collected for that stage are weighted and aggregated with their corresponding dynamic weights to obtain a quantitative score that comprehensively reflects the overall stress resistance of the variety at that growth stage. This process condenses multidimensional and heterogeneous indicator data into a representative single value, laying the foundation for subsequent time-series analysis. Second, "constructing a stress resistance state transition model to describe the evolution path of stress resistance states between stages and outputting a state transition stability score" means that each growth stage is considered as a different "state" on a time-series chain, and the calculated stress resistance score for each stage characterizes the variety's performance level in each "state." Based on the score sequence of all stages, a mathematical model (i.e., a state transition model) is constructed to characterize the correlation and transformation patterns between these states. This model can analyze how stress resistance performance influences the next stage (evolutionary path). For example, does excellent cold resistance in the seedling stage more likely lead to strong disease resistance in the budding stage? Simultaneously, the model can quantify the "stationarity" of this evolution, outputting a "state transition stationarity score" to assess whether the variety's stress resistance is stable and consistent throughout its growth period, or whether it exhibits significant fluctuations. This quantitative score reveals the temporal stability of stress resistance trait development.

[0062] For example, let's continue using the five stages (S1 to S5) of winter rapeseed. First, calculate the stage resistance score: For the S3 budding stage, assume its index system includes "sclerotinia rot incidence rate" (weight 0.4), "stem strength" (weight 0.35), and "number of days exceeding the precipitation standard" (weight 0.25). Obtain the measured data of a certain variety in the S3 stage: incidence rate 10% (normalized score 80), stem strength 8.5 N (score 85), precipitation exceeding the standard for 3 days (score 70). Then, the resistance score of this variety in the S3 stage, R_S3 = 80*0.4 + 85*0.35 + 70*0.25 = 78.25. Similarly, calculate the scores for S1, S2, S4, and S5 to obtain a score sequence, for example, [65, 72, 78, 85, 80]. Secondly, a state transition model is constructed and a score is output: the five-stage score sequence is input into a pre-defined Hidden Markov Model (HMM). Through learning, the model may reveal the following evolutionary path: varieties with high scores in S2 (overwintering stage) are significantly more likely to have high scores in S3 (budding stage), indicating that good overwintering performance is beneficial to subsequent growth. Simultaneously, the model analyzes the fluctuation of the entire score sequence: the sequence steadily rises from a moderate level (65, 72) to a peak (85) and then slightly declines (80), with relatively small overall fluctuations and a stable trend. Based on this, the model calculates a high "state transition stability score" (e.g., 8.5 out of 10), indicating that the variety's resistance develops in a coordinated manner throughout its entire growth period without obvious vulnerable breaks.

[0063] By combining the stage-specific stress resistance scores at each growth stage with the state transition stability scores, a dynamic stress resistance index for the target crop throughout its growth process is generated.

[0064] As mentioned above, this step is the final output and decision integration stage of this method. Its core lies in organically integrating the key information from the two dimensions obtained in the preliminary analysis—the static performance of each stage and the dynamic relationship between stages—to form a single, comprehensive, and time-series-rich quantitative evaluation index. First, "integrating the stage resistance scores of each growth stage" means integrating the scores of all growth stages, rather than simply summing them. This usually involves weighted averaging of the stage scores, with weights reflecting the relative importance of different growth stages throughout the entire growth period (possibly related to the "stage importance coefficient" used in calculating the dynamic weights of the aforementioned index, or it can be set independently). This ensures that the excellent or weak performance of key stages such as flowering is appropriately reflected in the final index. Second, "integrating the state transition stationarity score" refers to incorporating the stationarity score, which characterizes the quality of dynamic evolution, as an independent and important correction or combination term into the final calculation. Its purpose is to ensure that the final index not only reflects the variety's resistance at each "point" (each stage) but also whether the "line" (evolutionary process) connecting these "points" is smooth and coordinated. A high overall index should mean that the variety not only performs well at each stage, but also that its resilience development is robust, without drastic fluctuations or obvious weak points. The final "Dynamic Resilience Index" (DARI) thus becomes a high-level evaluation indicator that combines the dual attributes of "performance strength" and "development quality".

[0065] For example, continuing from the previous example of winter rapeseed, assume that the stress resistance scores for five stages have been calculated: S1=65, S2=72, S3=78, S4=85, and S5=80. Simultaneously, the state transition stability score output by the state transition model is 8.5 (out of 10). Now, generate the Dynamic Stress Resistance Index (DARI).

[0066] First, assign aggregate weights to the scores for each stage. Since the flowering and pod-setting stages (S3, S4) are crucial for final yield formation, they can be assigned higher weights (e.g., S3: 0.25, S4: 0.30), while other stages have lower weights (e.g., S1: 0.10, S2: 0.15, S5: 0.20). Calculate the weighted sum of the stage scores: 65*0.10 + 72*0.15 + 78*0.25 + 85*0.30 + 80*0.20 = 78.7.

[0067] Secondly, this weighted sum is combined with the state transition stationarity score. A feasible method is to set an adjustment coefficient λ for the stationarity score (e.g., λ=0.1), proportionally converting the stationarity score into a correction value for the weighted sum. Then, DARI = weighted sum of stage scores + λ * state transition stationarity score = 78.7 + 0.1 * 8.5 = 79.55. This value is the dynamic stress resistance index of this winter rapeseed variety throughout its entire life cycle. It not only includes the performance of each stage (78.7), but also receives a positive bonus (0.85) due to the stable evolution process (8.5), ultimately resulting in a more comprehensive evaluation.

[0068] According to one embodiment of this application, the dynamic weight of each evaluation index within a given stage is calculated using a time-series entropy weighting method that incorporates stage importance coefficients. Specifically, the calculation is performed using the following formula:

[0069]

[0070] in, For the first The first growth stage The dynamic weights of the evaluation indicators The entropy weight is calculated based on indicator data. The first [stage] determined based on the historical frequency of disaster occurrence and the duration of each stage The importance coefficient of each growth stage. and For the first Harmonization coefficients for each growth stage.

[0071] As mentioned above, the time-series entropy weighting method, which incorporates stage importance coefficients, clearly defines the specific process of calculating the dynamic weights of evaluation indicators within each stage using the given mathematical formula. The structure of this formula embodies the core idea of ​​combining objective data-driven weighting with prior knowledge guidance.

[0072] In the formula, This represents the dynamic weight of the i-th evaluation index in the j-th growth stage. The weight is calculated using two summation terms.

[0073] First item, , representing the weighted components calculated based on objective data. This is called entropy weight, and its calculation depends on an index. Historical or concurrent observational data sequences. Entropy weight is an objective weighting method based on the theory of information entropy. Its basic principle is that the greater the difference between different samples (such as different years or different varieties) of a certain indicator, that is, the higher the degree of dispersion, the greater the amount of information it can provide, and therefore it should be given greater weight in the comprehensive evaluation. It is the harmonic coefficient for stage j, used to adjust the proportion of the objective entropy weight component in the final dynamic weight of that stage.

[0074] The second item, , representing the weighted components determined based on prior agronomic knowledge. This is the stage importance coefficient for the j-th growth stage. This coefficient is not derived from real-time data, but is predetermined by comprehensively considering factors such as the historical frequency of disasters occurring in this stage (e.g., the probability and intensity of frost, drought, and pests and diseases) and the physiological duration of this stage within the entire growth period. A stage with frequent and severe disasters, or one that is crucial to the final yield of the crop, will be assigned a higher stage importance coefficient. It is another harmonic coefficient for stage j, used to adjust the contribution of this subjective knowledge component to the final weight.

[0075] By setting and The value of [the value] allows for a flexible balance between the influence of objective statistical laws and subjective experience in weight allocation. Ultimately, for any indicator within any growth stage, its dynamic weight [is determined]. All are determined by entropy weight, which reflects the information value of the data itself, and stage coefficient, which reflects the overall importance of the stage. They are then integrated through harmonic coefficients to form a more scientific and complex indicator weighting method that fits the complexity of agricultural production.

[0076] According to one embodiment of this application, the stress resistance state transition model is a hidden Markov model or a state-space model; the evolutionary path is used to identify the stress-vulnerable stage of the target crop during its growth period.

[0077] As mentioned above, the resilience state transition model is either a Hidden Markov Model (HMM) or a State Space Model (SSM). Both of these models are classic probabilistic graphical models for processing time series data, suitable for describing how the internal states of a system, which cannot be directly observed, evolve with time series (corresponding to crop growth stage series in this case).

[0078] Specifically, the model treats each growth stage as a time node, and the "stage resistance score" calculated for that stage is used as the observed value for that node. The "state" that the model focuses on is the crop's intrinsic and more fundamental level or pattern of resistance. Although this state cannot be directly measured, it is reflected through the observed scores of each stage. Hidden Markov Models (HMMs) model the stochastic transition process between resistance states at each stage by defining a set of hidden states, transition probabilities between states, and emission probabilities from states to observed values. State-space models, on the other hand, link the dynamic evolution of the system's internal state with external observations through a state equation and an observation equation, allowing for more flexible incorporation of system noise and observation errors.

[0079] The evolutionary path is used to identify the vulnerable stages of the target crop during its growth period. It refers to the ability to trace or predict the trajectory (evolutionary path) of stress resistance status throughout the entire growth period by analyzing the fitted state transition model. The identification of vulnerable stages is primarily based on the following principle: if model analysis indicates that the system state is highly susceptible to transitioning to a low-stress-resistance state at a specific growth stage, or if the observed score at that stage is extremely sensitive to state changes, then this stage can be identified as a "vulnerable stage." For example, the model might reveal an abnormally high probability of transitioning from a "high-stress-resistance state" to a "low-stress-resistance state" in the state transition probability matrix from stage A to stage B; in this case, the starting point of stage B might be identified as a vulnerable point. By analyzing the evolutionary path, specific growth periods where stress resistance is most likely to decline or pose a critical threat to overall stress resistance stability can be identified, thus providing precise decision-making basis for targeted field management.

[0080] According to one embodiment of this application, the formula for calculating the dynamic stress resistance index throughout the entire process is as follows:

[0081]

[0082] in, The dynamic resilience index throughout the entire process, This represents the total number of growth stages. For the first The overall weight of each growth stage For the first Stage-specific stress resistance scores for each growth stage Score the state transition stationarity. These are the preset transfer weight coefficients.

[0083] As mentioned above, the formula for generating the Dynamic Resilience Index (DARI) clearly defines the composition and calculation method of the index. The design of this formula embodies the core idea of ​​comprehensively integrating "phased static performance" with "inter-phase dynamic evolution quality".

[0084] The formula consists of two summation terms that together output a single quantitative evaluation value, DARI.

[0085] First item, This is a weighted summary of the static stress resistance performance across all growth stages (totaling n). Wherein:

[0086] The stage resistance score, calculated for the j-th growth stage, comprehensively reflects the level of stress resistance of the target crop at that specific stage.

[0087] This represents the overall weight assigned to the j-th growth stage. This weight is used to distinguish the differences in the contribution of different growth stages to the final overall stress resistance assessment. For example, the flowering and pod-setting stage, which is more critical to yield formation, may be assigned a higher weight.

[0088] By weighted summing of the scores for all stages, this calculation yields a basic score that reflects the overall strength of resilience throughout the entire reproductive period.

[0089] The second item, This aims to incorporate the dynamic characteristics of resilience evolving over time into the final evaluation. Specifically:

[0090] It is the state transition stability score previously output by the state transition model, which quantifies the coordination and stability of variety resistance during the transition between different growth stages.

[0091] λ is a preset transition weight coefficient used to adjust the proportion of the dynamic stationarity score in the final DARI index. By adjusting the value of λ, the influence of evolutionary process quality on the final evaluation result can be controlled.

[0092] Ultimately, the DARI index is the sum of the two values ​​mentioned above. This calculation formula ensures that the final evaluation index not only covers the crop's performance scores at each discrete growth stage, but also incorporates the stationarity assessment of the dynamic evolution of its stress resistance traits throughout the entire growth period, thus providing a more comprehensive and time-series insightful quantitative indicator of overall stress resistance.

[0093] According to one embodiment of this application, it also includes:

[0094] Based on the dynamic resilience index and the evolution path output by the resilience state transition model, a visual chart containing resilience evolution curves and / or stage comparison radar charts is generated.

[0095] As mentioned above, the system is based on the calculated stage stress resistance scores for each growth stage ( The graph generation module creates two main types of visualization charts, along with the resilient state evolution paths revealed by the state transition model.

[0096] Generate a stress resistance evolution curve: This curve is a two-dimensional line graph or curve. The horizontal axis represents the various growth stages arranged chronologically (e.g., S1, S2, … Sn), and the vertical axis represents the stress resistance score for that stage. The system uses the scores of each stage as data points, plots them sequentially on the graph, and connects them to form a trend line that visually displays the fluctuations in stress resistance levels as the reproductive process progresses. Combined with evolutionary path information output from the state transition model (such as state transition probabilities or predicted trends), the graph can be used to further highlight the stress-vulnerable stages identified by the model (e.g., highlighting stages where the score drops sharply or is at a low point), allowing users to clearly understand the dynamic process of changes in the variety's stress resistance strength and key weaknesses.

[0097] Growth Stage Comparison Radar Chart: This is a multi-dimensional chart. Each "axis" represents a specific growth stage or a core stress resistance trait within that stage (such as cold resistance or lodging resistance). The length or scale of each axis is normalized based on the score of the corresponding stage or trait. Connecting the score points of the target variety on each axis forms a polygonal region. The shape, size, and symmetry of this polygon visually reflect the differences and balance in the variety's stress resistance at different stages or for different traits. By overlaying and comparing radar charts of multiple varieties, the relative advantages and disadvantages of different varieties at specific stages or for specific stress resistance traits can be quickly identified.

[0098] These visualizations are not simply data transfers, but rather in-depth graphical representations based on model analysis results. They combine temporal dynamics (evolutionary curves) throughout the entire reproductive period with staged trait profiles (radar charts), providing users with multi-faceted insights from macro trends to micro structures, greatly enhancing the interpretability and decision support value of the dynamic resilience assessment results.

[0099] A second aspect of this application provides a dynamic stress resistance assessment system for oilseed crops based on growth stage division, comprising:

[0100] The data acquisition module is used to acquire field measured data, environmental monitoring data and disaster early warning information of the target crop at each preset growth stage throughout its entire growth period. The preset growth stages are predefined according to the crop's physiological cycle.

[0101] The indicator system construction module is used to construct an indicator system for evaluating the stress resistance of each growth stage based on the set of dominant stress resistance traits corresponding to each growth stage.

[0102] The weight calculation module is used to calculate the dynamic weight of each evaluation index within each growth stage using the time-series entropy weight method that incorporates the stage importance coefficient.

[0103] The stage scoring module is used to calculate the stage stress resistance score for each growth stage based on the indicator data and their dynamic weights within each stage.

[0104] The state transition analysis module is used to construct a stress resistance state transition model based on the stage stress resistance scores of all growth stages, to describe the evolution path of the stress resistance state between each stage, and to output the state transition stability score.

[0105] The index integration module is used to combine the stage stress resistance scores of each growth stage and the state transition stability scores to generate the full-process dynamic stress resistance index of the target crop.

[0106] According to one embodiment of this application, the weight calculation module is specifically used to calculate dynamic weights using the following formula:

[0107]

[0108] in, For the first The first growth stage The dynamic weights of the evaluation indicators The entropy weight is calculated based on indicator data. The first [stage] determined based on the historical frequency of disaster occurrence and the duration of each stage The importance coefficient of each growth stage. and For the first Harmonization coefficients for each growth stage.

[0109] According to one embodiment of this application, the model constructed by the state transition analysis module is a hidden Markov model or a state-space model; the evolution path is used to identify the stress-resistance and vulnerability stages of the target crop during its growth period.

[0110] According to one embodiment of this application, the calculation formula for the full-process dynamic resilience index generated by the index integration module is as follows:

[0111]

[0112] in, The dynamic resilience index throughout the entire process, This represents the total number of growth stages. For the first The overall weight of each growth stage For the first Stage-specific stress resistance scores for each growth stage Score the state transition stationarity. These are the preset transfer weight coefficients.

[0113] According to one embodiment of this application, it also includes:

[0114] The visualization module is used to generate visualization charts containing resilience evolution curves and / or stage comparison radar charts based on the evolution path output by the full-process dynamic resilience index and the resilience state transition model.

[0115] Example 2

[0116] 1. Delineation of growth stages

[0117] Based on the physiological cycle and agronomic management practices of winter rapeseed (Brassica napus L.), its entire growth period is pre-divided into five key growth stages:

[0118] S1 Seedling stage: From emergence to overwintering, the main growth stage is vegetative growth.

[0119] S2 overwintering period: from the start of overwintering to the return to green, experiencing low temperature stress.

[0120] S3 Budding stage: From the greening stage to the initial flowering stage, vegetative growth and reproductive growth proceed simultaneously.

[0121] S4 Flowering period: From the first bloom to the last bloom, the critical period for pollination and fruit setting.

[0122] S5 Maturity period: from the end of flowering to mechanical harvesting, grain filling and dehydration.

[0123] 2. Phased Data Acquisition

[0124] For each of the above stages, multi-source data will be collected synchronously at a test base in a major rapeseed producing area within a production cycle (e.g., 2023-2024).

[0125] Field measurement data: At the end of each stage, measurements were taken manually or through a sensor network. For example, the "overwintering survival rate" was measured after the S2 overwintering period; the "sclerotinia rot disease rate" was investigated during the S4 flowering period; and the "stem bending resistance" and "field lodging area ratio" were measured during the S5 maturity period.

[0126] Environmental monitoring data: Through small field weather stations, the "daily average temperature", "daily minimum temperature", "cumulative precipitation", "average wind speed" and other data are continuously recorded at various stages.

[0127] Disaster early warning information: Access the local agricultural meteorological service system to obtain official early warnings issued in each stage, such as the "low temperature freezing damage warning" for stage S2 and the "meteorological risk forecast of precipitation process and disease occurrence" for stages S3-S4.

[0128] 3. Construct a phased evaluation index system and calculate dynamic weights.

[0129] The dominant stress resistance traits were identified for each stage, and a corresponding set of indicators was constructed. Taking the S2 overwintering period and the S4 flowering period as examples:

[0130] Stage S2: The dominant trait is cold resistance. The indicator system includes: overwintering survival rate (measured), extreme low temperature value within the stage (environmental), and number of days with low temperature warnings (disaster). The time-series entropy weight method is used to calculate the entropy weight of each indicator based on data from the past five years of Stage S2. Combining the high StageWeight (e.g., 0.8) set for Stage S2 due to its frequent frost damage and its critical importance to yield, and the harmonic coefficients (α=0.6, β=0.4), the dynamic weight of each indicator in Stage S2 is calculated (e.g., overwintering survival rate has the highest weight).

[0131] Stage S4: The dominant traits are disease resistance and lodging resistance. The indicator system includes: sclerotinia rot incidence (measured), cumulative rainfall (environmental), number of strong wind warnings (disaster), and stem strength (measured). Using the same method, considering the criticality of the S4 flowering period to the final yield, the dynamic weights of each indicator in this stage are calculated.

[0132] 4. Calculation of resilience score

[0133] After normalizing the measured values ​​of the indicators collected during this production cycle, the values ​​are weighted and summed with the corresponding dynamic weights calculated in step 3 to calculate the stress resistance scores of variety A in the five stages (assuming a percentage system):

[0134] R_S1 = 75, R_S2 = 82, R_S3 = 78, R_S4 = 70, R_S5 = 85.

[0135] This sequence reflects the performance fluctuations of variety A at various stages, with S4 scoring lower during the flowering period.

[0136] 5. Constructing a state transition model and identifying vulnerable phases

[0137] The scoring sequence [75, 82, 78, 70, 85] from the five stages mentioned above was used as the observation sequence and input into a pre-defined Hidden Markov Model (HMM) for training and inference. Model analysis shows that:

[0138] Evolutionary path: Variety A's stress resistance was relatively stable in the early stage (S1-S3), but it entered a "low stress resistance state" with a high probability when transitioning from S3 to S4, and then recovered when transitioning from S4 to S5.

[0139] Identifying the Vulnerable Phase: The S4 flowering stage was identified by the model as a vulnerable phase. Its low score was not only due to its poor performance in its own indicators, but also because the model revealed that it was at a state transition point where it was prone to resilience decline.

[0140] Output stationarity score: Based on the volatility of the entire score sequence and the severity of state transitions, the model calculates TransitionStability = 72 (out of 100), indicating that the overall evolutionary stationarity is generally average and there are obvious fluctuations.

[0141] 6. Generate the Dynamic Resilience Index (DARI) throughout the entire process.

[0142] Set the overall weight W for each stage (reflecting the importance of the stage, such as higher weights for S2 and S4), and the transfer weight coefficient λ = 0.15.

[0143] calculate:

[0144] Total score for this stage = 75*0.1 + 82*0.2 + 78*0.2 + 70*0.3 + 85*0.2 = 77.4

[0145] Dynamic process bonus = 0.15 * 72 = 10.8

[0146] DARI = 77.4 + 10.8 = 88.2

[0147] 7. Visualization and Decision Support

[0148] The system automatically generates a visual report:

[0149] The resilience evolution curve: A curve connects the score points of the five stages, and marks the "vulnerable stage" at S4, visually showing the "funnel-shaped" decline in score during the S4 period.

[0150] Stage Comparison Radar Chart: Using five stages as five axes, the scoring outline of Variety A is drawn, clearly showing its obvious dip on the S4 axis.

[0151] Recommendations: The report summarizes that variety A exhibits good overall resistance (DARI=88.2), but its flowering period (S4) is a significantly vulnerable period, susceptible to diseases and adverse weather conditions, leading to large fluctuations in resistance. It is recommended to strengthen disease prevention (e.g., spraying fungicides) and field management (e.g., ditch clearing and drainage) during the S4 stage of this variety to stabilize its resistance and improve its stability throughout its growth period.

[0152] For any parts not mentioned in this application, existing technologies may be used or referenced.

[0153] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0154] The above description is merely an embodiment of this application and is not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. An oil crop dynamic stress resistance evaluation method based on growth stage division, characterized in that, The application comprises the following steps: Obtaining field measured data, environmental monitoring data and disaster warning information of the target crop at each preset growth stage in the whole growth period, wherein each preset growth stage is pre-determined according to the physiological cycle of the crop; Based on the dominant stress resistance trait set corresponding to each growth stage, an evaluation index system for each stage is constructed, and for each growth stage, the dynamic weight of each evaluation index in the stage is calculated by using a time sequence entropy weight method fused with a stage importance coefficient; According to the index data and the dynamic weight in each stage, the stage stress resistance score of each growth stage is calculated, and based on the stage stress resistance scores of all growth stages, a stress resistance state transition model is constructed to describe the evolution path of the stress resistance state between stages and output a state transition stability score; the stress resistance state transition model is a hidden Markov model or a state space model; the evolution path is used to identify the stress resistance fragile stage of the target crop in the growth period; The whole process dynamic stress resistance index of the target crop is generated by integrating the stage stress resistance scores of each growth stage and the state transition stability score.

2. The method of claim 1, wherein, The dynamic weight of each evaluation index in the stage is calculated by using a time sequence entropy weight method fused with a stage importance coefficient, and the dynamic weight is calculated by the following formula: in, For the first The first growth stage The dynamic weights of the evaluation indicators The entropy weight is calculated based on indicator data. The first [stage] determined based on the historical frequency of disaster occurrence and the duration of each stage The importance coefficient of each growth stage. and For the first Harmonization coefficients for each growth stage.

3. The method of claim 1, wherein, The calculation formula of the whole process dynamic stress resistance index is: wherein, is a total dynamic stress resistance index, is a total number of growth stages, is a comprehensive weight of the growth stage, is a stage stress resistance score of the growth stage, is the state transition stability score, is a preset transition weight coefficient.

4. The method according to any one of claims 1 to 3, characterized in that, The application further comprises the following steps: According to the whole process dynamic stress resistance index and the evolution path output by the stress resistance state transition model, a visual chart including a stress resistance evolution curve and / or a stage comparison radar chart is generated.

5. An oil crop dynamic stress resistance evaluation system based on growth stage division, characterized in that, The application comprises the following steps: A data acquisition module is configured to obtain field measured data, environmental monitoring data and disaster warning information of the target crop at each preset growth stage in the whole growth period, wherein each preset growth stage is pre-determined according to the physiological cycle of the crop; An index system construction module is configured to construct an evaluation index system for each stage based on the dominant stress resistance trait set corresponding to each growth stage; A weight calculation module is configured to calculate the dynamic weight of each evaluation index in the stage by using a time sequence entropy weight method fused with a stage importance coefficient for each growth stage; A stage scoring module is configured to calculate the stage stress resistance score of each growth stage according to the index data and the dynamic weight in each stage; A state transition analysis module is configured to construct a stress resistance state transition model based on the stage stress resistance scores of all growth stages to describe the evolution path of the stress resistance state between stages and output a state transition stability score; the stress resistance state transition model is a hidden Markov model or a state space model; the evolution path is used to identify the stress resistance fragile stage of the target crop in the growth period; An index integration module is configured to generate the whole process dynamic stress resistance index of the target crop by integrating the stage stress resistance scores of each growth stage and the state transition stability score.

6. The system of claim 5, wherein, The weight calculation module is specifically configured to calculate the dynamic weight by the following formula: in, For the first The first growth stage The dynamic weights of the evaluation indicators The entropy weight is calculated based on indicator data. The first [stage] determined based on the historical frequency of disaster occurrence and the duration of each stage The importance coefficient of each growth stage. and For the first Harmonization coefficients for each growth stage.

7. The system of claim 5, wherein, The model constructed by the state transition analysis module is a hidden Markov model or a state space model; the evolution path is used to identify the stress resistance fragile stage of the target crop in the growth period.

8. The system of claim 5, wherein, The calculation formula of the whole process dynamic stress resistance index generated by the index integration module is: wherein, is the total number of growth stages, is the total number of growth stages, is the comprehensive weight of the th growth stage, is the stage stress resistance score of the th growth stage, is the state transition stability score, is a preset transition weight coefficient.

9. The system according to any of claims 5-8, characterized in that, The application further comprises the following steps: a visualization module configured to generate a visualization chart comprising a stress evolution curve and / or a phase-contrast radar chart based on the full-range dynamic stress exponent and the evolution path output by the stress state transition model.

Images