A vegetation data optimization method and system for forestry environmental analysis
By constructing an intelligent agent for environmental parameters and vegetation prediction, and combining it with dynamic confidence sequence optimization of vegetation data, the problems of breeding decision lag and bias in traditional methods are solved, achieving efficient and accurate vegetation data optimization and improving the adaptability and stability of forestry breeding.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- XIDE COUNTY FORESTRY & GRASSLAND BUREAU
- Filing Date
- 2026-03-20
- Publication Date
- 2026-06-23
AI Technical Summary
Traditional vegetation data optimization methods rely on historical data and experience, making it difficult to predict future environmental changes. This leads to delayed and biased breeding decisions and an inability to adapt to the uncertainties of climate change.
An environmental parameter prediction agent is constructed to predict future environmental sequences. A dynamic environmental confidence sequence is introduced, and a vegetation prediction agent is combined to evaluate vegetation performance. The optimal vegetation data is selected through iterative optimization, and machine learning technology is used for data correction and matching degree calculation.
It significantly improves the adaptability and robustness of breeding decisions, realizes the dynamic closed-loop coupling of environmental prediction and breeding decisions, enhances breeding effectiveness and stability, reduces environmental adaptability risks, and promotes the development of forestry breeding towards intelligence and precision.
Smart Images
Figure CN121880321B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of vegetation data optimization technology, specifically to a method and system for optimizing vegetation data in forestry environment analysis. Background Technology
[0002] Forestry breeding, as a crucial foundation of forestry production, directly relates to the growth performance, stress resistance, and ecological adaptability of trees, playing a key role in improving the quality and efficiency of forestry resources. Traditional vegetation data optimization methods largely rely on human experience for evaluation and selection, often analyzing data based on historical climate data or current environmental conditions, lacking the ability to systematically predict and respond to future environmental changes. With intensifying climate change, environmental factors such as temperature and precipitation are exhibiting greater volatility and uncertainty, making static or empirical data analysis methods ill-suited to actual production needs and potentially leading to poor future performance of selected vegetation data. Summary of the Invention
[0003] This application provides a method and system for optimizing vegetation data in forestry environmental analysis, which addresses the technical problem that the optimization effect of vegetation data in the prior art is poor and cannot adapt to the environment.
[0004] In view of the above problems, this application provides a method and system for optimizing vegetation data in forestry environmental analysis.
[0005] In a first aspect, this application provides a method for optimizing vegetation data in forestry environmental analysis, the method comprising:
[0006] The system acquires historical environmental parameter sequences for the target forestry area within a preset historical time range, predicts environmental parameters for multiple future moments, obtains future environmental parameter sequences, and configures environmental confidence sequences based on the time intervals between multiple future moments.
[0007] Obtain the vegetation selection database to be selected, randomly select the first vegetation data, perform performance prediction, and obtain the first performance parameter. The prediction is performed using a vegetation prediction agent.
[0008] Based on the test environment parameters of different sample vegetation data and sample performance parameters in the vegetation prediction intelligent body, the environmental confidence sequence is corrected and calculated to obtain the corrected environmental confidence sequence.
[0009] Based on the first performance parameter, the future environmental parameter sequence, and the corrected environmental confidence sequence, the first vegetation adaptation value is obtained through analysis, and the vegetation data is further iterated and optimized to obtain the optimal vegetation data.
[0010] Secondly, this application provides a vegetation data optimization system for forestry environmental analysis, comprising:
[0011] The environmental parameter acquisition module is used to acquire the historical environmental parameter sequence of the target forestry area within a preset historical time range, predict the environmental parameters at multiple future moments, obtain the future environmental parameter sequence, and configure the environmental confidence sequence according to the time interval of multiple future moments.
[0012] The vegetation prediction module is used to obtain the vegetation selection database to be selected, randomly select the first vegetation data, perform performance prediction, and obtain the first performance parameter. The prediction is performed by a vegetation prediction agent.
[0013] The confidence acquisition module is used to perform correction calculations on the environmental confidence sequence based on the test environment parameters of different sample vegetation data and sample performance parameters in the vegetation prediction intelligent body, and obtain the corrected environmental confidence sequence.
[0014] The vegetation data optimization module is used to analyze and obtain the first vegetation adaptation value based on the first performance parameter, the future environmental parameter sequence, and the corrected environmental confidence sequence, and to continue iteratively optimizing the vegetation data to obtain the optimal vegetation data.
[0015] One or more technical solutions provided in this application have at least the following technical effects or advantages:
[0016] This application proposes a vegetation data optimization method and system for forestry environmental analysis. It constructs an environmental parameter prediction agent to accurately predict environmental sequences over multiple time periods. An environmental confidence sequence dynamically configured with time intervals is introduced to quantify prediction uncertainty. Simultaneously, the vegetation prediction agent pre-evaluates the performance of candidate vegetation data. The initial confidence is then corrected by analyzing the similarity between test environmental parameters in the agent's training data and the future predicted environment, thus forming a corrected environmental confidence sequence that better reflects the actual risk distribution. Finally, the vegetation fitness value is calculated comprehensively based on performance parameters, the future environment, and the corrected confidence. An iterative optimization mechanism is used to select the globally optimal vegetation data, significantly improving the adaptability and robustness of breeding decisions to future complex and variable environments. Compared with traditional methods, the technical solution provided in this application significantly overcomes the decision-making lag and bias caused by relying on historical static data or simple environmental extrapolation. It realizes the dynamic closed-loop coupling of environmental prediction and breeding decision-making, effectively enhances the stability and reliability of breeding programs under uncertain climatic conditions, and achieves the technical effects of improving breeding results, reducing environmental adaptability risks, and promoting the development of forestry breeding towards intelligence and precision. Attached Figure Description
[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0018] Figure 1 This is a flowchart illustrating a vegetation data optimization method for forestry environmental analysis provided in an embodiment of this application.
[0019] Figure 2 This is a schematic diagram of the structure of a vegetation data optimization system for forestry environment analysis provided in an embodiment of this application.
[0020] The components represented by each number in the attached diagram are explained below:
[0021] 100 - Environmental parameter acquisition module, 200 - Vegetation prediction module, 300 - Confidence acquisition module, 400 - Vegetation data optimization module. Detailed Implementation
[0022] This application provides a method and system for optimizing vegetation data in forestry environmental analysis, which addresses the technical problem that existing vegetation data optimization methods are ineffective and cannot adapt to the environment.
[0023] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.
[0024] It should be noted that the terms "comprising" and "having" are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or modules that are not explicitly listed or that are inherent to these processes, methods, products, or devices.
[0025] Example 1, as Figure 1 As shown, this application provides a method for optimizing vegetation data in forestry environmental analysis, wherein the method includes:
[0026] S10: Obtain the historical environmental parameter sequence of the target forestry area within a preset historical time range, predict the environmental parameters at multiple future moments, obtain the future environmental parameter sequence, and configure the environmental confidence sequence based on the time interval of multiple future moments.
[0027] In the field of vegetation data optimization, accurate prediction of future environmental conditions is the primary prerequisite for making scientific breeding decisions. How to construct precise sequences of future environmental parameters based on historical data, and how to assign a scientifically reasonable confidence weight to the prediction results at different time points to objectively characterize their predictive risks, have become technical challenges that constrain the foresight and robustness of breeding programs.
[0028] Step S10 in the method provided in this application embodiment includes:
[0029] The environmental parameters of the target forestry area within a preset historical time range are obtained and arranged chronologically to obtain a historical environmental parameter sequence, where each environmental parameter includes rainfall and temperature.
[0030] The historical environmental parameter sequence is input into the constructed environmental parameter prediction agent, and the output is the environmental parameters at multiple future moments, which are arranged in chronological order to obtain the future environmental parameter sequence.
[0031] The training steps for the environmental parameter prediction agent include:
[0032] Based on the environmental parameter monitoring records of the sample forestry area, a set of historical environmental parameter sequences of the samples was collected, and environmental parameters at multiple times after the historical environmental parameter sequences of different samples were collected to obtain a set of future environmental parameter sequences of the samples.
[0033] Using historical environmental parameter sequences as input and future environmental parameter sequences as output, an intelligent agent for predicting environmental parameters is constructed based on machine learning.
[0034] The environmental parameter prediction agent is trained using the set of historical environmental parameter sequences and the set of future environmental parameter sequences of the samples, and the training is completed after convergence.
[0035] Obtain multiple time intervals between future moments and the current moment. Based on these multiple time intervals, calculate and assign multiple environment confidence scores, forming an environment confidence score sequence. The magnitude of the time intervals and the magnitude of the environment confidence scores are negatively correlated. The formula for calculating the environment confidence score is as follows:
[0036] Environmental confidence level = 1 / (1 + time interval).
[0037] In this embodiment, environmental parameters of the target forestry area within a preset historical time range are obtained. For example, the preset historical time range is set to the past two years to make the data closer to the current situation. The historical environmental parameter sequence is obtained by arranging the parameters chronologically. Each environmental parameter includes rainfall in millimeters and temperature in degrees Celsius.
[0038] To construct an intelligent agent for predicting environmental parameters, firstly, based on the environmental parameter monitoring records of the sample forestry area, a set of historical environmental parameter sequences of the samples is collected, and then environmental parameters at multiple times after the historical environmental parameter sequences of different samples are collected, sorted and integrated in chronological order to obtain a set of future environmental parameter sequences of the samples.
[0039] Based on machine learning, an intelligent agent for predicting environmental parameters is constructed. For example, a three-layer structure is adopted, in which the input layer is used to receive the historical environmental parameter sequence, the hidden layer is an LSTM hidden layer containing 128 neural units, and the fully connected output layer is used to output the predicted environmental parameter sequence.
[0040] Using historical environmental parameter sequences as input and future environmental parameter sequences as output, an environmental parameter prediction agent is trained using a set of sample historical environmental parameter sequences and a set of sample future environmental parameter sequences. Alternatively, a supervised learning approach is used, with the sample historical environmental parameter sequences as input and the corresponding sample future environmental parameter sequences as the target output. During training, the Adam optimizer is used to minimize the mean squared error between the predicted output and the true future sequence. Training stops when the agent's loss value on the validation set no longer decreases significantly, indicating that the environmental parameter prediction agent has converged and training is complete.
[0041] The historical environmental parameter sequence is input into the constructed environmental parameter prediction agent, and the output is the environmental parameters at multiple future moments. The parameters are then arranged in chronological order to obtain the future environmental parameter sequence.
[0042] We obtain multiple time intervals between future moments and the current moment, and calculate multiple environmental confidence scores to form an environmental confidence score sequence. The size of the time interval is negatively correlated with the size of the environmental confidence score. For example, predicting day 1 in the future uses a time interval of 1; predicting day 30 uses a time interval of 30. Then, we assign confidence scores according to a preset rule: the smaller the time interval, the higher the confidence score; the larger the time interval, the lower the confidence score. Environmental confidence score = 1 ÷ (1 + time interval). According to this formula, the confidence score for day 1 in the future = 1 ÷ (1 + 1) = 0.5, and the confidence score for day 2 in the future = 1 ÷ (1 + 2) = 0.33. Arranging the confidence score values calculated for each prediction point in the next 30 days in sequence yields the environmental confidence score sequence.
[0043] By constructing an intelligent agent for predicting environmental parameters and performing deep learning and pattern mining on historical environmental parameter sequences, high-precision sequential prediction of environmental parameters at multiple future moments was achieved, generating a more comprehensive and coherent description of future environmental scenarios. An environmental confidence sequence matching the environmental prediction sequence was introduced and dynamically allocated based on the time interval between future and current moments, strictly adhering to the objective law that the uncertainty of long-term predictions is higher than that of short-term predictions. This mechanism transforms the originally vague concept of prediction reliability into a quantifiable confidence index, providing a risk measurement scale for subsequent breeding decisions. It significantly improves the scientific rigor of the entire vegetation adaptability assessment process and its ability to cope with uncertain futures from the source, laying a solid and reliable data foundation for subsequent steps.
[0044] S20: Obtain the vegetation selection database to be selected, randomly select the first vegetation data, perform performance prediction, and obtain the first performance parameter, wherein a vegetation prediction agent is used for prediction.
[0045] After obtaining information on future environmental predictions, the core task is to assess the potential performance of specific vegetation data under this environment. Traditional methods mostly rely on limited field trial data or rough inferences based on experience. This process is time-consuming, costly, and unable to cover a large number of possible combinations, severely limiting the efficiency and scope of breeding screening.
[0046] Step S20 in the method provided in this application embodiment includes:
[0047] Based on the vegetation types required for the target forestry area, a vegetation selection database is constructed using all breeding categories of the aforementioned vegetation types.
[0048] Randomly select the first vegetation data from the vegetation selection database;
[0049] The first vegetation data is input into the vegetation prediction agent, and the first performance parameter is output. The first performance parameter includes drought resistance parameter and heat resistance parameter. The vegetation prediction agent is constructed using the sample performance parameters of vegetation test under sample test environment parameters from multiple sets of sample vegetation data.
[0050] The construction steps of the vegetation prediction agent include:
[0051] Based on historical vegetation test data, a set of sample vegetation data, a set of sample test environment parameters, and a set of sample performance parameters obtained from the test were collected.
[0052] Using vegetation data as input and the first performance parameter as output, a vegetation prediction agent is constructed based on machine learning.
[0053] The vegetation prediction agent is trained using the sample vegetation data set and sample performance parameter set as training data, and the training is completed after convergence.
[0054] In this embodiment, a vegetation selection database is constructed using all breeding categories of the target forestry area, based on the required vegetation type. Vegetation types may include, but are not limited to, different tree species such as *Pinus sylvestris* and *Pinus slashensis*, as well as different cultivated varieties of the same species, such as "disease-resistant *Pinus sylvestris*" and "fast-growing *Pinus slashensis*". All these available breeding categories are then compiled to form a vegetation selection database containing all possible options.
[0055] The first vegetation data is randomly selected from the vegetation selection database as the current processing object. For example, two varieties, "Pinus sylvestris" and "fast-growing slash pine", are randomly selected from the database and combined together to form the first vegetation data in order to predict the performance after hybridization breeding.
[0056] A vegetation prediction agent is constructed, exemplarily based on a random forest regression algorithm. The input layer receives vegetation data, the hidden layer of the random forest exemplarily uses 100 decision trees, and the output layer outputs the first performance parameters for prediction. These first performance parameters include drought resistance and heat resistance parameters. The drought resistance parameter, measured in millimeters, represents the precipitation range required by the vegetation data; precipitation below the minimum or above the maximum of this range may lead to poor growth. Similarly, the heat resistance parameter, measured in temperature, represents the temperature required by the vegetation data.
[0057] Based on historical vegetation test data, a set of sample vegetation data, a set of sample test environment parameters, and a set of sample performance parameters obtained from the test were collected.
[0058] A vegetation prediction agent is trained using a sample vegetation dataset and a sample performance parameter dataset, learning the mapping relationship between vegetation data and performance parameters. The training data is divided into a training set and a validation set at a ratio of 8:2. During training, the agent's performance is tested on the validation set. The agent is considered to have converged and training is complete when its prediction error on the validation set no longer decreases significantly.
[0059] Input the first vegetation data into the vegetation prediction agent and output the first performance parameters.
[0060] By introducing a pre-trained vegetation prediction agent to address the problem of inefficient vegetation data selection, this agent can simulate and learn the complex mapping relationship between combinations and performance in historical vegetation test data. By inputting random vegetation data, the agent can output predicted values for multiple key performance parameters, replacing traditional time-consuming physical testing. This enables rapid initial screening of a large amount of candidate vegetation data, greatly expanding the optimization search space and improving the efficiency of vegetation data selection.
[0061] S30: Based on the test environment parameters of different sample vegetation data and sample performance parameters in the vegetation prediction intelligent body, the environmental confidence sequence is corrected and calculated to obtain the corrected environmental confidence sequence.
[0062] Different vegetation data rely on different historical testing environments for their performance predictions. If the future prediction environment differs significantly from the historical testing environment of a particular combination, even if the performance requirements for the vegetation data are similar, the actual reliability of the predictions made by the agent trained on that historical data will be low, potentially leading to significant deviations in the selection of breeding programs.
[0063] Step S30 in the method provided in this application embodiment includes:
[0064] Obtain the sample test environment parameters corresponding to the first vegetation data within the training data of the vegetation prediction agent;
[0065] Calculate the similarity between the sample test environment parameters and multiple future environment parameters within the future environment parameter sequence to obtain a test environment similarity sequence;
[0066] Based on the test environment similarity sequence, a confidence correction coefficient sequence is calculated, and the environment confidence sequence is corrected to obtain a corrected environment confidence sequence.
[0067] In this embodiment of the application, the sample test environment parameters corresponding to the first vegetation data within the training data of the vegetation prediction agent are obtained.
[0068] The similarity between the sample test environment parameters and multiple future environment parameters within the future environment parameter sequence is calculated to obtain a test environment similarity sequence. Specifically, the similarity between rainfall and temperature at the same time in the sample test environment parameter sequence and the future environment parameter sequence is calculated separately. For example, rainfall similarity = 1 - |sample test rainfall - future rainfall| ÷ [(sample test rainfall + future rainfall) ÷ 2]. The same method is used to calculate temperature similarity. The mean of rainfall similarity and temperature similarity is taken as the test environment similarity. Multiple test environment similarities are integrated in chronological order to obtain the test environment similarity sequence.
[0069] Based on the test environment similarity sequence, a confidence correction coefficient sequence is calculated, and the environment confidence sequence is then corrected to obtain a corrected environment confidence sequence. For example, the test environment similarity at each time point in the test environment similarity sequence is used as the confidence correction coefficient. That is, the higher the test environment similarity, the more similar the test environment is to the predicted future environment, resulting in higher prediction accuracy and a smaller correction magnitude; conversely, the lower the similarity, the larger the correction magnitude. Using the confidence correction coefficient sequence, the environment confidence sequence is corrected by multiplying the values at corresponding positions in the two sequences, i.e., corrected environment confidence = environment confidence × confidence correction coefficient. The final corrected environment confidence sequence not only considers the uncertainty caused by the time distance but also incorporates information about the similarity between historical test environments and future predicted environments.
[0070] By mining the sample test environment parameters in the training data of the vegetation prediction agent and comparing their similarity with the future environment prediction sequence, personalized dynamic correction of the environment confidence sequence is achieved. By calculating the similarity between the future environment and the historical test environment, a correction coefficient is generated to adjust the confidence level. This makes the corrected environment confidence sequence more accurately reflect the measurement of the uncertainty in predicting the current vegetation data, allowing subsequent fitness calculations to be based on a more realistic and accurate risk assessment.
[0071] S40: Based on the first performance parameter, the future environmental parameter sequence, and the corrected environmental confidence sequence, analyze and obtain the first vegetation adaptation value, and continue to iterate and optimize the vegetation data to obtain the optimal vegetation data.
[0072] Traditional methods often simply make a binary comparison between performance and environmental requirements, failing to fully consider the risk weight of environmental predictions and the complex matching relationship between performance and environment. This may lead to the final decision falling into local optima or failing to scientifically balance short-term performance and long-term risks.
[0073] Step S40 in the method provided in this application embodiment includes:
[0074] Calculate the degree of matching between the first performance parameter and each future environmental parameter in the future environmental parameter sequence to obtain multiple first environmental matching degrees;
[0075] Calculate the similarity between multiple future environmental parameters within the future environmental parameter sequence and the sample test environmental parameters to obtain multiple first environmental similarities;
[0076] Multiple first-environment fitness values are calculated based on multiple first-environment matching degrees and multiple first-environment similarities.
[0077] According to the modified environmental confidence sequence, the multiple first environmental fitness values are weighted and calculated to obtain the first vegetation fitness value;
[0078] Continue to randomly select vegetation data for iterative optimization until the optimization converges and the optimal vegetation data with the maximum vegetation fitness value is obtained.
[0079] In this embodiment, the matching degree between the first performance parameter and each future environmental parameter in the future environmental parameter sequence is calculated to obtain multiple first environmental matching degrees. Specifically, the matching degree between the first performance parameter and the future environmental parameter is calculated. For example, if a future environmental parameter in the future environmental parameter sequence is a precipitation of 150 mm and a temperature of 35℃, the corresponding drought resistance parameter is 140 mm and the heat resistance parameter is 30℃, then the first environmental matching degree = (drought resistance parameter similarity + heat resistance parameter similarity) ÷ 2, where the drought resistance parameter similarity = [1 - |precipitation - drought resistance parameter| ÷ [(precipitation + drought resistance parameter) ÷ 2] = 0.93, and the heat resistance parameter similarity = [1 - |temperature - heat resistance parameter| ÷ [(temperature + heat resistance parameter) ÷ 2] = 0.84, then the first environmental matching degree = (0.93 + 0.84) ÷ 2 = 0.88.
[0080] The similarity between multiple future environmental parameters and sample test environmental parameters within a future environmental parameter sequence is calculated to obtain multiple first environmental similarities. Specifically, the similarity between rainfall and temperature at the same time in the sample test environmental parameter sequence and the future environmental parameter sequence is calculated separately. For example, rainfall similarity = 1 - |sample test rainfall - future rainfall| ÷ [(sample test rainfall + future rainfall) ÷ 2]. The same method is used to calculate temperature similarity. Finally, the mean of rainfall similarity and temperature similarity is calculated to obtain the first environmental similarity.
[0081] Multiple first-environment fitness values are calculated based on multiple first-environment matching degrees and multiple first-environment similarities. The first-environment fitness value = (first-environment matching degree + first-environment similarity) ÷ 2. The first-environment fitness value represents the degree of adaptability of the current breeding program to the future environment. The larger the first-environment fitness value, the better the adaptability of the current breeding program to the future environment.
[0082] The first vegetation fitness value is obtained by weighting multiple first environmental fitness values according to the modified environmental confidence sequence. Specifically, the first vegetation fitness value = ∑[first environmental fitness value × (modified environmental confidence ÷ sum of modified environmental confidence sequences)], which comprehensively considers the breeding adaptability under future environmental changes. The larger the first vegetation fitness value, the better the adaptability of the first vegetation data in the future dynamically changing environment.
[0083] Continue to randomly select vegetation data and perform iterative optimization until the optimization converges and the optimal vegetation data with the maximum vegetation fitness value is obtained.
[0084] By organically integrating the matching degree between the primary performance parameter and the future environmental sequence, the similarity between the future environment and the historical test environment, and the modified environmental confidence sequence representing risk weights, a comprehensive and scientific single evaluation index is obtained. This index simultaneously considers expected performance, empirical reliability, and risk preference. Iterative optimization then converges to the global optimum, resulting in a fully validated and optimized optimal breeding scheme that can adapt to uncertain future environments and has controllable risks. This significantly improves the overall quality and efficiency of the acquired forestry vegetation data.
[0085] Example 2, as Figure 2 As shown, based on the same inventive concept as the vegetation data optimization method for forestry environment analysis provided in Embodiment 1, this embodiment of the invention also provides a vegetation data optimization system for forestry environment analysis, comprising:
[0086] The environmental parameter acquisition module 100 is used to acquire the historical environmental parameter sequence of the target forestry area within a historical preset time range, predict the environmental parameters at multiple future times, obtain the future environmental parameter sequence, and configure the environmental confidence sequence according to the time interval of multiple future times.
[0087] The vegetation prediction module 200 is used to obtain a vegetation selection database for vegetation to be selected, randomly select the first vegetation data, perform performance prediction, and obtain the first performance parameter. The prediction is performed by a vegetation prediction agent.
[0088] The confidence acquisition module 300 is used to perform correction calculations on the environmental confidence sequence based on the test environment parameters of different sample vegetation data and sample performance parameters in the vegetation prediction intelligent body, and obtain a corrected environmental confidence sequence.
[0089] The vegetation data optimization module 400 is used to analyze and obtain the first vegetation adaptation value based on the first performance parameter, the future environmental parameter sequence and the corrected environmental confidence sequence, and to continue iterative vegetation data optimization to obtain the optimal vegetation data.
[0090] In one embodiment, the environmental parameter acquisition module 100 is further configured to:
[0091] The environmental parameters of the target forestry area within a preset historical time range are obtained and arranged chronologically to obtain a historical environmental parameter sequence, where each environmental parameter includes rainfall and temperature.
[0092] The historical environmental parameter sequence is input into the constructed environmental parameter prediction agent, and the output is the environmental parameters at multiple future moments, which are arranged in chronological order to obtain the future environmental parameter sequence.
[0093] The training steps for the environmental parameter prediction agent include:
[0094] Based on the environmental parameter monitoring records of the sample forestry area, a set of historical environmental parameter sequences of the samples was collected, and environmental parameters at multiple times after the historical environmental parameter sequences of different samples were collected to obtain a set of future environmental parameter sequences of the samples.
[0095] Using historical environmental parameter sequences as input and future environmental parameter sequences as output, an intelligent agent for predicting environmental parameters is constructed based on machine learning.
[0096] The environmental parameter prediction agent is trained using the set of historical environmental parameter sequences and the set of future environmental parameter sequences of the samples, and the training is completed after convergence.
[0097] Obtain multiple time intervals between future moments and the current moment. Based on these multiple time intervals, calculate and assign multiple environment confidence scores, forming an environment confidence score sequence. The magnitude of the time intervals and the magnitude of the environment confidence scores are negatively correlated. The formula for calculating the environment confidence score is as follows:
[0098] Environmental confidence level = 1 / (1 + time interval).
[0099] In one embodiment, the vegetation prediction module 200 is further configured to:
[0100] Based on the vegetation types required for the target forestry area, a vegetation selection database is constructed using all breeding categories of the aforementioned vegetation types.
[0101] Randomly select the first vegetation data from the vegetation selection database;
[0102] The first vegetation data is input into the vegetation prediction agent, and the first performance parameter is output. The first performance parameter includes drought resistance parameter and heat resistance parameter. The vegetation prediction agent is constructed using the sample performance parameters of vegetation test under sample test environment parameters from multiple sets of sample vegetation data.
[0103] The construction steps of the vegetation prediction agent include:
[0104] Based on historical vegetation test data, a set of sample vegetation data, a set of sample test environment parameters, and a set of sample performance parameters obtained from the test were collected.
[0105] Using vegetation data as input and the first performance parameter as output, a vegetation prediction agent is constructed based on machine learning.
[0106] The vegetation prediction agent is trained using the sample vegetation data set and sample performance parameter set as training data, and the training is completed after convergence.
[0107] In one embodiment, the confidence acquisition module 300 is further configured to:
[0108] Obtain the sample test environment parameters corresponding to the first vegetation data within the training data of the vegetation prediction agent;
[0109] Calculate the similarity between the sample test environment parameters and multiple future environment parameters within the future environment parameter sequence to obtain a test environment similarity sequence;
[0110] Based on the test environment similarity sequence, a confidence correction coefficient sequence is calculated, and the environment confidence sequence is corrected to obtain a corrected environment confidence sequence.
[0111] In one embodiment, the vegetation data optimization module 400 is further configured to:
[0112] Calculate the degree of matching between the first performance parameter and each future environmental parameter in the future environmental parameter sequence to obtain multiple first environmental matching degrees;
[0113] Calculate the similarity between multiple future environmental parameters within the future environmental parameter sequence and the sample test environmental parameters to obtain multiple first environmental similarities;
[0114] Multiple first-environment fitness values are calculated based on multiple first-environment matching degrees and multiple first-environment similarities.
[0115] According to the modified environmental confidence sequence, the multiple first environmental fitness values are weighted and calculated to obtain the first vegetation fitness value;
[0116] Continue to randomly select vegetation data for iterative optimization until the optimization converges and the optimal vegetation data with the maximum vegetation fitness value is obtained.
[0117] In summary, the embodiments of this application have at least the following technical effects:
[0118] This application proposes a vegetation data optimization method and system for forestry environmental analysis. It constructs an environmental parameter prediction agent to accurately predict environmental sequences over multiple time periods. An environmental confidence sequence dynamically configured with time intervals is introduced to quantify prediction uncertainty. Simultaneously, the vegetation prediction agent pre-evaluates the performance of candidate vegetation data. The initial confidence is then corrected by analyzing the similarity between test environmental parameters in the agent's training data and the future predicted environment, thus forming a corrected environmental confidence sequence that better reflects the actual risk distribution. Finally, the vegetation fitness value is calculated comprehensively based on performance parameters, the future environment, and the corrected confidence. An iterative optimization mechanism is used to select the globally optimal vegetation data, significantly improving the adaptability and robustness of breeding decisions to future complex and variable environments. Compared with traditional methods, the technical solution provided in this application significantly overcomes the decision-making lag and bias caused by relying on historical static data or simple environmental extrapolation. It realizes the dynamic closed-loop coupling of environmental prediction and breeding decision-making, effectively enhances the stability and reliability of breeding programs under uncertain climatic conditions, and achieves the technical effects of improving breeding results, reducing environmental adaptability risks, and promoting the development of forestry breeding towards intelligence and precision.
[0119] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.
[0120] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.
[0121] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and modifications fall within the scope of this application and its equivalents, this application intends to include such modifications and modifications.
Claims
1. A method for optimizing vegetation data in forestry environmental analysis, characterized in that, The method includes: The system acquires historical environmental parameter sequences for the target forestry area within a preset historical time range, predicts environmental parameters for multiple future moments, obtains future environmental parameter sequences, and configures environmental confidence sequences based on the time intervals between multiple future moments. Obtain the vegetation selection database to be selected, randomly select the first vegetation data, perform performance prediction, and obtain the first performance parameter. The prediction is performed using a vegetation prediction agent. Based on the test environment parameters of different sample vegetation data and sample performance parameters in the vegetation prediction intelligent body, the environmental confidence sequence is corrected and calculated to obtain the corrected environmental confidence sequence. Based on the first performance parameter, the future environmental parameter sequence, and the corrected environmental confidence sequence, the first vegetation adaptation value is obtained through analysis, and iterative optimization is continued to obtain the optimal vegetation data. Obtain historical environmental parameter sequences for the target forestry area within a preset historical time range, predict environmental parameters for multiple future time points, obtain future environmental parameter sequences, and configure environmental confidence sequences based on the time intervals between multiple future time points, including: The environmental parameters of the target forestry area within a preset historical time range are obtained and arranged chronologically to obtain a historical environmental parameter sequence, where each environmental parameter includes rainfall and temperature. The historical environmental parameter sequence is input into the constructed environmental parameter prediction agent, and the output is the environmental parameters at multiple future moments, which are arranged in chronological order to obtain the future environmental parameter sequence. Obtain multiple time intervals between future moments and the current moment. Based on these multiple time intervals, calculate and assign multiple environment confidence scores, forming an environment confidence score sequence. The magnitude of the time intervals and the magnitude of the environment confidence scores are negatively correlated. The formula for calculating the environment confidence score is as follows: Environmental confidence level = 1 / (1 + time interval); Based on the test environment parameters of different sample vegetation data and sample performance parameters within the vegetation prediction intelligent system, the environmental confidence sequence is corrected and calculated to obtain a corrected environmental confidence sequence, including: Obtain the sample test environment parameters corresponding to the first vegetation data within the training data of the vegetation prediction agent; Calculate the similarity between the sample test environment parameters and multiple future environment parameters within the future environment parameter sequence to obtain a test environment similarity sequence; Based on the test environment similarity sequence, a confidence correction coefficient sequence is calculated, and the environment confidence sequence is corrected to obtain a corrected environment confidence sequence.
2. The vegetation data optimization method for forestry environmental analysis according to claim 1, characterized in that, The training steps for the environmental parameter prediction agent include: Based on the environmental parameter monitoring records of the sample forestry area, a set of historical environmental parameter sequences of the samples was collected, and environmental parameters at multiple times after the historical environmental parameter sequences of different samples were collected to obtain a set of future environmental parameter sequences of the samples. Using historical environmental parameter sequences as input and future environmental parameter sequences as output, an intelligent agent for predicting environmental parameters is constructed based on machine learning. The environmental parameter prediction agent is trained using the set of historical environmental parameter sequences and the set of future environmental parameter sequences of the samples, and the training is completed after convergence.
3. The vegetation data optimization method for forestry environmental analysis according to claim 1, characterized in that, Obtain the vegetation selection database for the vegetation to be selected, randomly select the first vegetation data, perform performance prediction, and obtain the first performance parameters, including: Based on the vegetation types required for the target forestry area, a vegetation selection database is constructed using all breeding categories of the aforementioned vegetation types. Randomly select the first vegetation data from the vegetation selection database; The first vegetation data is input into the vegetation prediction agent, and the first performance parameters are output. The first performance parameters include drought resistance parameters and heat resistance parameters. The vegetation prediction agent is constructed using the sample performance parameters of vegetation tests conducted under sample test environment parameters from multiple sets of sample vegetation data.
4. The vegetation data optimization method for forestry environmental analysis according to claim 3, characterized in that, The steps for constructing the vegetation prediction agent include: Based on historical vegetation test data, a set of sample vegetation data, a set of sample test environment parameters, and a set of sample performance parameters obtained from the test were collected. Using vegetation data as input and performance parameters as output, a vegetation prediction agent is constructed based on machine learning. The vegetation prediction agent is trained using the sample vegetation data set and sample performance parameter set as training data, and the training is completed after convergence.
5. The vegetation data optimization method for forestry environmental analysis according to claim 1, characterized in that, Based on the first performance parameter, the future environmental parameter sequence, and the corrected environmental confidence sequence, the first vegetation adaptation value is obtained through analysis, and iterative optimization is continued to obtain the optimal vegetation data, including: Calculate the degree of matching between the first performance parameter and each future environmental parameter in the future environmental parameter sequence to obtain multiple first environmental matching degrees; Calculate the similarity between multiple future environmental parameters within the future environmental parameter sequence and the sample test environmental parameters to obtain multiple first environmental similarities; Multiple first-environment fitness values are calculated based on multiple first-environment matching degrees and multiple first-environment similarities. According to the modified environmental confidence sequence, the multiple first environmental fitness values are weighted and calculated to obtain the first vegetation fitness value; Continue to randomly select vegetation data for iterative optimization until the optimization converges and the optimal vegetation data with the maximum vegetation fitness value is obtained.
6. A vegetation data optimization system for forestry environmental analysis, characterized in that, A vegetation data optimization method for implementing a forestry environmental analysis according to any one of claims 1-5, the system comprising: The environmental parameter acquisition module is used to acquire the historical environmental parameter sequence of the target forestry area within a preset historical time range, predict the environmental parameters at multiple future moments, obtain the future environmental parameter sequence, and configure the environmental confidence sequence according to the time interval of multiple future moments. The vegetation prediction module is used to obtain the vegetation selection database to be selected, randomly select the first vegetation data, perform performance prediction, and obtain the first performance parameter. The prediction is performed by a vegetation prediction agent. The confidence acquisition module is used to perform correction calculations on the environmental confidence sequence based on the test environment parameters of different sample vegetation data and sample performance parameters in the vegetation prediction intelligent body, and obtain the corrected environmental confidence sequence. The vegetation data optimization module is used to analyze and obtain the first vegetation adaptation value based on the first performance parameter, the future environmental parameter sequence, and the corrected environmental confidence sequence, and to continue iterative optimization to obtain the optimal vegetation data.