Soil erosion prevention, moisture conservation and fertilization full life cycle benefit evaluation and optimization system

Abstract

The present invention discloses a soil erosion prevention, moisture conservation and fertilization full life cycle benefit evaluation optimization system, which relates to the field of soil optimization management and control technology. The system includes: a full life cycle data stream acquisition module, which is used to obtain multi-party soil erosion prevention, moisture conservation and fertilization full life cycle data streams; a multi-dimensional distributed training fusion module, which is used to build a global soil full life cycle benefit evaluation model set; a model verification optimization module, which is used to collect the local soil erosion prevention, moisture conservation and fertilization full life cycle data streams in the target area, and optimize the local soil full life cycle benefit evaluation model; a soil benefit optimization management and control module, which is used to perform soil benefit optimization management and control. The present invention solves the technical problems of insufficient multi-dimensional data integration and analysis, weak model generalization ability, and low benefit evaluation accuracy in traditional soil management in the prior art, and achieves the technical effect of improving the benefit evaluation accuracy through the fusion analysis of multi-source data and the local optimization of the model.

Description

Technical Field

[0001] The present invention relates to the technical field of soil optimization and control, and in particular to a full life cycle benefit evaluation and optimization system for soil erosion prevention, moisture conservation and fertilization. Background Art

[0002] Soil erosion prevention, soil moisture conservation and soil fertility improvement technology is an important part of modern agricultural management, involving a series of soil protection and improvement measures to maintain or improve soil productivity and ensure the sustainable development of agriculture. With global climate change, soil degradation and increasing pressure on agricultural production, how to effectively prevent soil erosion, maintain soil moisture and improve soil fertility has become a core challenge in agricultural management.

[0003] However, soil erosion prevention, soil moisture conservation and soil fertility measures involve multiple data types, which come from different sources and formats, making them difficult to integrate and analyze in a unified manner. In addition, traditional soil benefit evaluation models can often only process data from a single data source or limited dimensions, which limits the applicability and prediction accuracy of the model. Applications in specific areas may also be affected by factors such as the regional environment and soil characteristics, resulting in inaccurate evaluation results. Summary of the invention

[0004] The present application provides a full life cycle benefit evaluation and optimization system for soil erosion prevention, moisture conservation and fertilization, which is used to solve the technical problems in the existing technology of traditional soil management, such as insufficient multi-dimensional data integration and analysis, weak model generalization ability, and low benefit evaluation accuracy.

[0005] The present application provides a soil erosion prevention, moisture conservation and fertilization full life cycle benefit evaluation optimization system, the system comprising: a full life cycle data stream acquisition module, the full life cycle data stream acquisition module is used to obtain multiple soil erosion prevention, moisture conservation and fertilization full life cycle data streams through data mining technology, the multiple soil erosion prevention, moisture conservation and fertilization full life cycle data streams include soil basic data, environmental data, water carbon footprint data and agricultural production data; a multi-dimensional distributed training fusion module, the multi-dimensional distributed training fusion module is used to obtain a set of soil benefit evaluation indicators, based on the soil benefit evaluation indicator set, the multi-dimensional distributed training fusion of the multiple soil erosion prevention, moisture conservation and fertilization full life cycle data streams is carried out to construct a global soil full life cycle. A set of periodic benefit evaluation models; a model verification and optimization module, the model verification and optimization module is used to collect and obtain the local soil erosion prevention, moisture conservation and fertilization full life cycle data stream in the target area, and based on the global soil full life cycle benefit evaluation model set, the model verification and optimization of the local soil erosion prevention, moisture conservation and fertilization full life cycle data stream is performed to obtain the local soil full life cycle benefit evaluation model; a soil benefit optimization management and control module, the soil benefit optimization management and control module is used to perform key factor analysis on the target area based on the local soil full life cycle benefit evaluation model, obtain a set of key influencing factors of the soil full life cycle, and perform soil benefit optimization management and control through the set of key influencing factors of the soil full life cycle.

[0006] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0007] The soil erosion prevention, moisture conservation and fertilization full life cycle benefit evaluation optimization system provided in the present application relates to the field of soil optimization and control technology. It obtains multi-party soil erosion prevention, moisture conservation and fertilization full life cycle data streams through data mining technology, conducts multi-dimensional distributed training fusion, constructs a global soil full life cycle benefit evaluation model set, and performs local optimization to obtain a local soil full life cycle benefit evaluation model, conducts key factor analysis on the target area, obtains a set of key influencing factors of the soil full life cycle for soil benefit optimization and control, solves the technical problems of insufficient multi-dimensional data integration and analysis, weak model generalization ability, and low benefit evaluation accuracy in traditional soil management in the prior art, realizes the improvement of benefit evaluation accuracy through the fusion analysis of multi-source data and the local optimization of the model, screens out the optimal mode for the synergistic improvement of crop comprehensive benefits, and achieves the technical effect of synergistic optimization of erosion prevention, moisture conservation, fertilization and production capacity improvement. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0009] Figure 1 A schematic diagram of the structure of a soil erosion prevention, moisture conservation and fertilization full life cycle benefit evaluation optimization system provided in an embodiment of the present application;

[0010] Figure 2 A schematic diagram of a process for constructing a global soil life cycle benefit evaluation model set in a soil erosion prevention, moisture conservation and fertilization life cycle benefit evaluation optimization system provided in an embodiment of the present application;

[0011] Figure 3 A schematic diagram of a process for obtaining a set of key influencing factors for the entire life cycle of soil in the entire life cycle benefit evaluation and optimization system for soil erosion prevention, moisture conservation and fertilization provided in an embodiment of the present application.

[0012] Explanation of the accompanying drawings: full life cycle data stream acquisition module 11, multi-dimensional distributed training fusion module 12, model verification and optimization module 13, soil benefit optimization and control module 14. DETAILED DESCRIPTION

[0013] The present application provides a full life cycle benefit evaluation and optimization system for soil erosion prevention, moisture conservation and fertilization, which is used to solve the technical problems in the existing technology of traditional soil management, such as insufficient multi-dimensional data integration and analysis, weak model generalization ability, and low benefit evaluation accuracy.

[0014] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0015] It should be noted that the terms "first", "second", etc. in the specification of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, 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 clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products, or devices.

[0016] Example

[0017] like Figure 1 As shown, the present application provides a soil erosion prevention, moisture conservation and fertilization full life cycle benefit evaluation optimization system, the system comprising:

[0018] The full life cycle data stream acquisition module 11 is used to obtain multiple soil erosion prevention, moisture conservation and fertilization full life cycle data streams through data mining technology. The multiple soil erosion prevention, moisture conservation and fertilization full life cycle data streams include soil basic data, environmental data, water carbon footprint data and agricultural production data.

[0019] Specifically, the full life cycle data stream acquisition module 11 of the present application is used to systematically acquire and integrate multi-party data streams through advanced data mining technology to support the full life cycle benefit evaluation of soil erosion prevention, moisture conservation and fertilization. In order to achieve this goal, the module first extensively consulted and analyzed domestic and foreign research results in the fields of soil erosion prevention, moisture conservation, fertilization technology, capacity improvement, full life cycle assessment (LCA), cost-benefit analysis, economic benefits, environmental benefits and comprehensive benefit evaluation. These research results provide a scientific basis and technical support for the collection and analysis of data streams.

[0020] In terms of data collection, Module 11 pays special attention to basic soil data, environmental data, water carbon footprint data, and agricultural production data. These data cover key information such as the physical and chemical properties of the soil, local climatic conditions, resource consumption of agricultural activities (such as water usage and carbon emissions), and crop yields. Basic soil data include soil type, texture, organic matter content, etc., which are important indicators for evaluating soil health and suitability. Environmental data involve climatic factors such as temperature, precipitation, wind speed, etc., which have a direct impact on soil and crop growth conditions. Water carbon footprint data represents the efficiency of water use and carbon emissions in the agricultural production process, and is an important component of measuring environmental benefits. Agricultural production data include crop type, yield, fertilization, etc., which are the basis for evaluating production capacity and economic benefits.

[0021] Through data mining technology, these multi-party data are systematically collected, integrated and stored in the database. The application of data mining technology can not only efficiently process large amounts of data, but also identify patterns and trends hidden in the data, providing reliable data support for subsequent model training. After completing the data collection, specific data from multiple research areas are selected and trained under different model frameworks. By comparing and analyzing the training effects of these models, the most suitable model framework can be determined, providing a scientific basis for the subsequent evaluation of soil erosion prevention, moisture conservation and fertilization benefits.

[0022] A multi-dimensional distributed training fusion module 12 is used to obtain a set of soil benefit evaluation indicators, and based on the soil benefit evaluation indicator set, multi-dimensional distributed training fusion is performed on the multi-party soil erosion prevention, moisture conservation and fertilization full life cycle data streams to construct a global soil full life cycle benefit evaluation model set.

[0023] Furthermore, the multi-dimensional distributed training fusion module 12 is also used to perform the following steps:

[0024] P21: Quantify the correlation and screen the indicators of the soil benefit evaluation indicator set to obtain a strongly correlated benefit evaluation indicator set; P22: Analyze and select the hydrological model framework according to the soil characteristic information of the target area to determine the benefit evaluation model training framework set; P23: Based on the strongly correlated benefit evaluation indicator set, perform multi-dimensional indicator classification on the multi-party soil erosion prevention, moisture conservation and fertilization full life cycle data streams to obtain a multi-party indicator soil full life cycle data stream set; P24: Based on the benefit evaluation model training framework set, perform distributed training and fusion on the multi-party indicator soil full life cycle data stream set to construct the global soil full life cycle benefit evaluation model set.

[0025] It should be understood that the multi-dimensional distributed training fusion module 12 of the present application is responsible for converting the complex soil benefit evaluation process into an operational mathematical model. The module first defines a comprehensive set of soil benefit evaluation indicators, which not only includes direct indicators of soil erosion prevention effect (such as erosion rate, erosion reduction, etc.), soil moisture retention capacity (such as soil moisture content, water retention capacity improvement, etc.) and soil fertility status (such as nutrient cycle efficiency improvement, crop yield and quality increase, etc.), but also comprehensively considers the economic benefits, environmental benefits, ecological benefits and social benefits of soil erosion prevention, moisture retention and fertilization measures throughout the life cycle. Through this comprehensive evaluation system, the module can fully and deeply understand the actual effects of different soil management measures.

[0026] Specifically, we first used statistical and machine learning methods to quantify and screen the correlation of the soil benefit evaluation index set. By calculating the correlation coefficient between the indicators and performing principal component analysis, we identified those indicators that have a significant impact on the benefits of the entire life cycle of the soil and are highly independent of each other, forming a set of strongly correlated benefit evaluation indicators. This process effectively reduced redundant information and improved the efficiency and accuracy of model training.

[0027] Furthermore, considering that the soil characteristics (such as soil type, topography, climate conditions, etc.) of different target areas vary greatly, the existing hydrological model frameworks are analyzed and selected based on these characteristic information. By comparing the applicability and prediction accuracy of different models under specific conditions, a series of benefit evaluation model training frameworks suitable for different regions are determined. These frameworks provide a solid theoretical foundation and technical support for subsequent distributed training.

[0028] Next, based on a set of strongly correlated benefit evaluation indicators, a multi-dimensional indicator classification of the data streams of the entire life cycle of multi-party soil erosion prevention, moisture conservation and fertilization was performed. Through pre-processing steps such as data cleaning, conversion and standardization, the original data stream was divided into multiple subsets corresponding to specific evaluation indicators to form a multi-party indicator soil life cycle data stream set. This process ensures the accuracy and consistency of the data and provides high-quality data input for subsequent training fusion.

[0029] Finally, using distributed computing technology, the multi-indicator soil life cycle data stream set is trained and integrated based on the benefit evaluation model training framework set. Through the application of parallel processing and optimization algorithms, the module can quickly build a series of global soil life cycle benefit evaluation model sets for different regions and evaluation dimensions. These models not only have high prediction accuracy and generalization ability, but also can provide decision makers with comprehensive and accurate soil management recommendations.

[0030] In summary, through the above steps, the multi-dimensional distributed training fusion module 12 successfully transforms the complex soil benefit evaluation problem into a set of operational mathematical models, providing strong support for the accurate evaluation and optimization of the benefits of soil erosion prevention, moisture conservation and fertilization throughout the life cycle.

[0031] Furthermore, step P21 of the embodiment of the present application further includes:

[0032] P21-1: Extract any one evaluation indicator from the soil benefit evaluation indicator set in turn; P21-2: Based on any one evaluation indicator, traverse and calculate in the multi-party soil erosion prevention, moisture conservation and fertilization full life cycle data stream to obtain a benefit indicator correlation coefficient set; P21-3: Weightedly calculate and integrate the benefit indicator correlation coefficient set according to the soil benefit evaluation indicator set to obtain a benefit indicator correlation degree set; P21-4: Set a preset correlation degree threshold, screen and match the benefit indicator correlation degree set based on the preset correlation degree threshold, and obtain the strongly correlated benefit evaluation indicator set.

[0033] Optionally, in order to more accurately screen out the evaluation indicators that have a significant impact on the benefits of the entire life cycle of soil, that is, the set of strongly correlated benefit evaluation indicators, first, the system will extract any evaluation indicator from the pre-defined set of soil benefit evaluation indicators in turn as the current analysis object. This process is automated, ensuring that each indicator can be fairly and comprehensively evaluated.

[0034] Then, based on the currently extracted evaluation indicators, traversal calculations are performed in the data streams of the entire life cycle of soil erosion prevention, moisture conservation and fertilization. For example, correlation analysis methods in statistics, such as the Pearson correlation coefficient and the Spearman rank correlation coefficient, are used to calculate the correlation coefficients between the current indicator and all other indicators. These correlation coefficients reflect the synchronization or trend of changes between different indicators and are an important basis for evaluating the degree of interdependence between indicators.

[0035] After obtaining the benefit index correlation coefficient set of each indicator, these coefficients are further weighted and integrated according to the soil benefit evaluation index set. This step takes into account the importance and influence of different indicators in the overall evaluation system, and calculates the comprehensive correlation of each indicator by assigning different weight values. The comprehensive correlation not only reflects the direct correlation between indicators, but also incorporates the relative importance of indicators in the entire evaluation system.

[0036] Finally, a preset correlation threshold is set, which is determined based on previous research, expert experience or statistical laws to distinguish between strongly correlated and weakly correlated indicators. The comprehensive correlation of each indicator is compared with the preset correlation threshold, and those indicators with a comprehensive correlation higher than the threshold are screened out to form a set of strongly correlated benefit evaluation indicators. These strongly correlated indicators will play a key role in subsequent model training because they can more accurately reflect the key influencing factors of the benefits of the entire life cycle of soil.

[0037] Further, such as Figure 2 As shown, step P24 of the embodiment of the present application also includes:

[0038] P24-1: Based on the benefit evaluation model training framework set, the multi-indicator soil life cycle data stream set is distributedly trained to obtain a multi-framework multi-indicator benefit evaluation model set; P24-2: According to the benefit indicator correlation set, the strongly correlated benefit evaluation indicator set is weighted to determine the benefit indicator weight factor set; P24-3: Based on the benefit indicator weight factor set, the multi-framework multi-indicator benefit evaluation model set is decision-fused to obtain a multi-framework multi-party benefit fusion model set; P24-4: According to the benefit evaluation model training framework set, the model parameters of the multi-framework multi-party benefit fusion model set are extracted and aggregated for training to construct the global soil life cycle benefit evaluation model set.

[0039] Specifically, in order to further improve the accuracy and generalization ability of the soil life cycle benefit assessment model, advanced technologies such as distributed training, decision fusion and federated learning can be used to deeply process and integrate multi-indicator soil life cycle data streams.

[0040] First, based on the benefit evaluation model training framework set, the multi-indicator soil life cycle data stream set is distributedly trained. This process fully utilizes the advantages of distributed computing, splits large-scale data streams into multiple subsets, and processes them in parallel on different computing nodes. Each node trains the corresponding data stream subset according to the specified benefit evaluation model training framework, and finally obtains a multi-framework multi-indicator benefit evaluation model set. These model sets contain multi-indicator evaluation models trained according to different training frameworks for multiple research areas, providing a wealth of candidate models for subsequent decision fusion.

[0041] Next, according to the benefit indicator correlation set, the strongly correlated benefit evaluation indicator set is weighted. By quantifying the importance of different indicators in the overall evaluation, a weight factor is assigned to each strongly correlated indicator to form a benefit indicator weight factor set. These weight factors reflect the contribution of different indicators to the benefits of the entire life cycle of soil, providing an important reference for subsequent model fusion.

[0042] Then, based on the set of benefit indicator weight factors, the decision fusion of the multi-frame and multi-party indicator benefit evaluation model set is carried out. The output results of multiple evaluation models can be comprehensively processed by using decision fusion algorithms such as weighted average and Bayesian fusion to obtain a set of multi-frame and multi-party benefit fusion models. These fusion models not only contain information on each research area, but also strengthen and weaken key indicators through the adjustment of weight factors, thereby improving the overall evaluation performance of the model.

[0043] Finally, according to the benefit evaluation model training framework set, the model parameters of the multi-framework multi-party benefit fusion model set are extracted and aggregated for training. This process draws on the idea of ​​federated learning and jointly trains the indicator fusion evaluation models from different research areas. Under the premise of protecting the data privacy of each research area, the model parameters are shared and optimized by exchanging the updated parameters of the model instead of the original data. Finally, a global soil life cycle benefit evaluation model set is constructed. These models integrate data source information from multiple research areas and have stronger generalization ability and accuracy. At the same time, since the data imbalance and data bias problems that may exist in a single data source are avoided, these models are also more robust and reliable, and can effectively avoid the occurrence of overfitting.

[0044] The model verification and optimization module 13 is used to collect and obtain the local soil erosion prevention, moisture conservation and fertilization full life cycle data stream in the target area, and based on the global soil full life cycle benefit evaluation model set, the model verification and optimization of the local soil erosion prevention, moisture conservation and fertilization full life cycle data stream is performed to obtain a local soil full life cycle benefit evaluation model.

[0045] Furthermore, the model verification optimization module 13 is also used to perform the following steps:

[0046] P31: Based on the global soil life cycle benefit evaluation model set, the local soil erosion protection, moisture conservation and fertilization life cycle data streams are respectively model verified and evaluated to obtain a multi-framework model performance evaluation parameter set; P32: According to the multi-framework model performance evaluation parameter set, the global soil life cycle benefit evaluation model set is optimized to obtain a target soil life cycle benefit evaluation model; P33: According to the matching of the target soil life cycle benefit evaluation model and the multi-framework model performance evaluation parameter set, the target model performance evaluation parameters are determined; P34: Based on the target model performance evaluation parameters, the target soil life cycle benefit evaluation model is iteratively optimized to obtain the local soil life cycle benefit evaluation model.

[0047] It should be understood that the model verification and optimization module 13 of the present application is used to collect and analyze the local soil erosion prevention, moisture conservation and fertilization full life cycle data stream in the target area, use the global model for verification and optimization, and localize the global soil full life cycle benefit evaluation model set into a soil benefit evaluation model for a specific target area.

[0048] Specifically, first collect and obtain the full life cycle data stream of local soil erosion prevention, moisture conservation and fertilization in the target area, which usually includes historical monitoring data, experimental data and expert experience knowledge in the area. Then, use each model in the global soil full life cycle benefit evaluation model set to verify and evaluate the local data stream. By comparing the difference between the model prediction results and the actual observation data, the performance evaluation parameters of each model are calculated, such as accuracy, recall rate, F1 score, mean square error, etc., to form a multi-framework model performance evaluation parameter set. This process provides a quantitative basis for subsequent model optimization.

[0049] Next, based on the multi-frame model performance evaluation parameter set, the effect optimization operation is further performed. This step usually involves the comprehensive ranking of performance parameters, weight allocation, and the application of decision-making algorithms (such as voting mechanism, weighted average, etc.). By comprehensively considering the advantages and disadvantages of different models, the target soil life cycle benefit evaluation model that best suits the soil characteristics of the target area is finally determined. This model not only shows high prediction accuracy in a specific area, but also can better reflect the soil ecological benefit characteristics of the area.

[0050] After determining the target soil life cycle benefit evaluation model, the module will further determine the performance evaluation parameters of the target model based on the matching of the model with the multi-frame model performance evaluation parameter set. These parameters not only reflect the performance of the model on the validation data set, but also provide an important reference benchmark for subsequent model iteration optimization.

[0051] Finally, the model validation and optimization module 13 iteratively optimizes the target soil life cycle benefit evaluation model based on the target model performance evaluation parameters. This process may include various optimization strategies such as fine-tuning of parameters, adjustment of model structure, and introduction of regularization terms. Through continuous trial and error and improvement, the prediction accuracy and generalization ability of the model are further improved, and finally a local soil life cycle benefit evaluation model suitable for the target area is constructed. This model can not only provide a scientific basis for local soil management and agricultural decision-making, but also provide valuable reference for soil benefit evaluation in other similar areas.

[0052] Furthermore, step P34 of the embodiment of the present application also includes:

[0053] P34-1: Optimize and analyze the performance evaluation parameters of the target model and determine the model optimization strategy; P34-2: Extract the model parameter information of the target soil life cycle benefit evaluation model, and the model parameter information includes model weights and model biases; P34-3: Perform strategy optimization on the model parameter information based on the model optimization strategy, and define a fitness loss function to perform gradient calculation on the model parameter information after strategy optimization to obtain model parameter gradient information; P34-4: Iteratively update the model parameter information based on the model parameter gradient information through a back propagation algorithm to obtain the local soil life cycle benefit evaluation model.

[0054] In a possible embodiment of the present application, in order to further improve the accuracy and generalization ability of the target soil life cycle benefit evaluation model, the specific process of model optimization and parameter adjustment may be, first, an in-depth analysis of the target model performance evaluation parameters is conducted to identify possible deviations and deficiencies in the model during the prediction process. This process may involve comparing the performance of the model on different data sets, or using methods such as sensitivity analysis to evaluate the impact of various parameters on model performance. Based on these analysis results, the model optimization strategy is determined, including but not limited to adjusting the model structure, optimizing algorithm selection, regularization parameter setting, etc. These strategies are designed to improve the performance of the model in a targeted manner to make it more adaptable to the soil characteristics of the target area.

[0055] Subsequently, the parameter extraction algorithm was used to extract the model parameter information of the target soil life cycle benefit assessment model, mainly including model weights and model biases. Model weights refer to the strength of the internal connections of the model, which are used to adjust the impact of input features on the output results. Model bias is an additional parameter used to balance the output of the model to make it closer to the true value. These parameters are the key information learned by the model during the training process, and they determine the prediction behavior of the model. By accurately obtaining these parameter information, basic data support is provided for subsequent strategy optimization.

[0056] Furthermore, based on the determined model optimization strategy, the model parameter information is optimized. This process may include reinitializing parameters, adjusting learning rates, introducing new optimization algorithms, etc. At the same time, in order to quantify the optimization effect and guide the subsequent optimization direction, a fitness loss function is defined, which is used to measure the difference between the model prediction results and the actual observed data. Then, the gradient descent and other optimization algorithms are used to calculate the gradient of the model parameter information after strategy optimization to obtain the model parameter gradient information. This information indicates the direction and size of parameter adjustment, which is the key to improving model performance.

[0057] Finally, the model parameters are iteratively updated using the model parameter gradient information through the back propagation algorithm. The back propagation algorithm is an effective neural network training algorithm that can reversely adjust the parameter values ​​in the network according to the gradient information of the loss function. In the iterative update process, the model parameter update steps are repeatedly performed to gradually optimize the model parameters. Through continuous iterative adjustments, the model parameters gradually approach the optimal solution, and the value of the loss function also decreases accordingly. When the loss function value reaches the preset threshold or the number of iterations reaches the preset upper limit, the iterative update process is stopped, and the model obtained at this time is the local soil full life cycle benefit evaluation model. After sufficient optimization and adjustment, the model can more accurately reflect the benefit characteristics of the soil in the target area throughout its life cycle, providing strong support for soil management and agricultural decision-making.

[0058] The soil benefit optimization management and control module 14 is used to perform key factor analysis on the target area based on the local soil full life cycle benefit evaluation model, obtain a set of key influencing factors of the soil full life cycle, and optimize soil benefit management through the set of key influencing factors of the soil full life cycle.

[0059] Specifically, the soil benefit optimization and control module 14 of the present application is responsible for conducting an in-depth analysis of the target area based on the local soil full life cycle benefit evaluation model to achieve optimization and control of soil benefits.

[0060] For example, first receive the local soil life cycle benefit assessment model as the basis for analysis. The model has been rigorously verified and optimized, and can accurately reflect the benefit characteristics of the soil in the target area throughout its life cycle. Subsequently, the model is used to conduct a key factor analysis of the target area. This step usually involves an in-depth interpretation of the various indicators output by the model, combining knowledge from multiple fields such as soil science, agronomy, and environmental science to identify the key factors that affect the benefits of the soil throughout its life cycle. These key factors may include soil type, climatic conditions, farming methods, fertilization strategies, water resource management, etc.

[0061] After obtaining a set of key factors affecting the entire life cycle of soil, further soil benefit optimization management strategies are formulated. This process requires comprehensive consideration of multiple factors such as the actual situation of the target area, policy orientation, economic benefits, and ecological and environmental needs. By constructing a multi-objective optimization model or using expert systems and other methods, a series of targeted management and control measures are proposed to improve soil quality, increase crop yields, reduce environmental pollution, and promote sustainable agricultural development, achieve precise management, maximize soil benefits, and reduce environmental burdens.

[0062] Further, such as Figure 3 As shown, the soil benefit optimization control module 14 is also used to perform the following steps:

[0063] P41: Obtain the soil erosion prevention, moisture conservation and fertilization data stream to be evaluated in the target area; P42: Evaluate the soil erosion prevention, moisture conservation and fertilization data stream to be evaluated based on the local soil full life cycle benefit evaluation model, and output the soil full life cycle benefit evaluation parameters; P43: Analyze the influencing factors of the soil benefit evaluation index set, and obtain the benefit index influencing factor set; P44: Construct a multi-objective benefit decision-making model based on the soil benefit evaluation index set and the benefit index influencing factor set; P45: Analyze the key factors of the soil full life cycle benefit evaluation parameters based on the multi-objective benefit decision-making model, and obtain the key influencing factor set of the soil full life cycle.

[0064] It should be understood that the specific process of obtaining a set of key factors affecting the entire life cycle of soil can be, first, obtaining the soil erosion prevention and soil moisture conservation and fertilization data stream to be evaluated from the monitoring system of the target area. These data streams contain real-time and historical information on soil moisture, nutrient content, erosion degree, crop growth status, etc., which are the basis for evaluating the benefits of the entire life cycle of soil. In order to ensure the accuracy and consistency of the data, these data streams are preprocessed, including data cleaning (removing noise and outliers), data conversion (unifying dimensions and formats), and data aggregation (integration by time or space dimensions).

[0065] Next, the pre-processed soil erosion prevention, moisture conservation and fertilization data stream to be evaluated was evaluated using the local soil life cycle benefit evaluation model that has been built and verified. Through model calculation, a series of soil life cycle benefit evaluation parameters are output, such as soil quality index, crop yield forecast, water resource utilization efficiency, environmental pollution risk, etc. These parameters fully reflect the benefit performance of soil under different management strategies.

[0066] After obtaining the soil benefit evaluation parameters, the influencing factors of the soil benefit evaluation index set are further analyzed. This process involves in-depth analysis of each indicator to identify the key factors affecting soil benefits. Through statistical analysis, causal reasoning or machine learning and other technical means, a correlation model between benefit indicators and soil erosion prevention, moisture conservation and fertilization technology production influencing factors is constructed to obtain a set of influencing factors of benefit indicators. These influencing factors may include soil type, fertilizer type and amount, irrigation method, crop variety, farming system, etc.

[0067] Furthermore, a multi-objective benefit decision model is constructed based on the set of soil benefit evaluation indicators and the set of influencing factors of benefit indicators. This model aims to comprehensively consider the trade-offs between multiple benefit indicators and the joint effects of different influencing factors on benefit indicators, and provide scientific decision support for soil management. Multi-objective decision models usually use multi-objective optimization algorithms (such as genetic algorithms, particle swarm algorithms, etc.) to solve the optimal soil management strategy by setting optimization goals and constraints.

[0068] Finally, the key factor analysis of the soil life cycle benefit evaluation parameters is carried out based on the multi-objective benefit decision model. Through model calculation and analysis, the key factor set with the greatest impact on the soil life cycle benefit is identified. These key factors are not only an important basis for formulating soil management optimization strategies, but also the focus of future research and improvement. According to the results of the key factor analysis, the module will formulate corresponding optimization strategies, including adjusting fertilization plans, improving irrigation technology, optimizing crop layout, strengthening soil erosion prevention, etc., to maximize soil benefits.

[0069] Furthermore, the soil benefit optimization control module 14 is also used to perform the following steps:

[0070] P46: Construct a soil erosion prevention, moisture conservation and fertilization strategy library, match the set of key influencing factors of the entire life cycle of the soil with the soil erosion prevention, moisture conservation and fertilization strategy library, and obtain multiple soil optimization strategy schemes; P47: Perform simulation evaluation based on the multiple soil optimization strategy schemes to obtain benefit evaluation information of multiple strategy schemes, optimize the multiple soil optimization strategy schemes through the benefit evaluation information of the multiple strategy schemes, and determine the target soil optimization strategy scheme; P48: Perform soil benefit optimization management and control on the target area based on the target soil optimization strategy scheme.

[0071] Optionally, the specific process of optimizing soil benefits management and control through the set of key influencing factors of the entire life cycle of the soil may be to construct a soil erosion prevention, moisture conservation and fertilization strategy library after identifying the key influencing factors of the entire life cycle of the soil. This strategy library is a database that integrates a variety of soil management technologies and practical experience, and contains optimization strategies for different soil types, climatic conditions, crop types and other situations. Through an intelligent matching algorithm, the set of key influencing factors of the entire life cycle of the soil is compared with the various strategies in the strategy library to screen out multiple soil optimization strategy solutions that best fit the current target area. This step can use big data analysis and machine learning technology to achieve personalized and precise strategy recommendations.

[0072] In order to evaluate the effects of different soil optimization strategies, simulation technology is used to test each solution in a virtual environment. The simulation process is based on mathematical models of soil physics, chemistry and biological processes, and can simulate changes in soil quality, crop growth, water resource utilization and other aspects after the implementation of different strategies. By collecting and analyzing the benefit evaluation information generated by the simulation, the module can quantitatively compare multiple soil optimization strategies, so as to select the target soil optimization strategy with the best benefits and the highest feasibility. It can not only improve the scientificity and accuracy of decision-making, but also avoid the risks and costs that may be brought about in actual operations.

[0073] After determining the target soil optimization strategy, Module 14 will guide and assist the target area in implementing the strategy. This includes developing a detailed implementation plan, allocating resources, training relevant personnel, and monitoring the progress of implementation. At the same time, the module will continue to monitor and evaluate soil benefits to ensure the effective implementation of the optimization strategy and the continuous improvement of soil benefits. If deviations or problems are found during the implementation process, the strategy will be adjusted in a timely manner or remedial measures will be taken to ensure the realization of the soil benefit optimization control objectives. This step reflects the initiative and flexibility of the module in soil management, and also provides strong support for its important role in sustainable agricultural development.

[0074] In summary, the embodiments of the present application have at least the following technical effects:

[0075] This application uses data mining technology to obtain multi-party soil erosion prevention, moisture conservation and fertilization full life cycle data streams, conducts multi-dimensional distributed training fusion, constructs a global soil full life cycle benefit evaluation model set, and performs local optimization to obtain a local soil full life cycle benefit evaluation model, conducts key factor analysis on the target area, and obtains a set of key influencing factors of the soil full life cycle for soil benefit optimization management and control.

[0076] The goal is to improve the accuracy of benefit evaluation through the fusion analysis of multi-source data and the local optimization of the model, screen out the optimal model for synergistic improvement of the comprehensive benefits of crops, and achieve the technical effect of synergistic optimization of erosion prevention, moisture conservation, fertilization and production capacity improvement.

[0077] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0078] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.

[0079] This specification and the drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.

Claims

1. The whole life cycle benefit evaluation and optimization system for soil erosion prevention, moisture conservation and fertilization is characterized by: The system comprises: A full life cycle data stream acquisition module, which is used to acquire multiple soil erosion prevention, moisture conservation and fertilization full life cycle data streams through data mining technology, and the multiple soil erosion prevention, moisture conservation and fertilization full life cycle data streams include soil basic data, environmental data, water carbon footprint data and agricultural production data; A multi-dimensional distributed training fusion module, which is used to obtain a set of soil benefit evaluation indicators, and based on the set of soil benefit evaluation indicators, multi-dimensional distributed training fusion is performed on the multi-party soil erosion prevention, moisture conservation and fertilization full life cycle data streams to build a global soil full life cycle benefit evaluation model set; A model verification and optimization module, wherein the model verification and optimization module is used to collect and obtain the local soil erosion prevention, moisture conservation and fertilization full life cycle data stream in the target area, and based on the global soil full life cycle benefit evaluation model set, respectively perform model verification and optimization on the local soil erosion prevention, moisture conservation and fertilization full life cycle data stream to obtain a local soil full life cycle benefit evaluation model; A soil benefit optimization management and control module, which is used to perform key factor analysis on the target area based on the local soil full life cycle benefit evaluation model, obtain a set of key influencing factors of the soil full life cycle, and perform soil benefit optimization management and control through the set of key influencing factors of the soil full life cycle; The local soil full life cycle benefit assessment model is obtained, including: Based on the global soil life cycle benefit evaluation model set, model validation and evaluation are performed on the local soil erosion prevention, moisture conservation and fertilization life cycle data streams to obtain a multi-framework model performance evaluation parameter set; According to the multi-frame model performance evaluation parameter set, the global soil life cycle benefit evaluation model set is optimized to obtain a target soil life cycle benefit evaluation model; Determine the target model performance evaluation parameters according to the matching of the target soil life cycle benefit evaluation model and the multi-framework model performance evaluation parameter set; Iteratively optimizing the target soil life cycle benefit evaluation model based on the target model performance evaluation parameters to obtain the local soil life cycle benefit evaluation model; Optimizing and analyzing the target model performance evaluation parameters to determine the model optimization strategy; Extracting and obtaining model parameter information of the target soil life cycle benefit assessment model, wherein the model parameter information includes model weight and model bias; Based on the model optimization strategy, the model parameter information is optimized, and at the same time, a fitness loss function is defined to perform gradient calculation on the model parameter information after the strategy optimization to obtain model parameter gradient information; The model parameter information is iteratively updated based on the model parameter gradient information through a back propagation algorithm to obtain the local soil full life cycle benefit evaluation model.

2. The soil erosion prevention, moisture conservation and fertilization full life cycle benefit evaluation and optimization system according to claim 1, characterized in that: The construction of a global soil life cycle benefit assessment model set includes: Performing correlation quantification and index screening on the soil benefit evaluation index set to obtain a strongly correlated benefit evaluation index set; Analyze and select the hydrological model framework according to the soil characteristic information of the target area to determine the benefit evaluation model training framework set; Based on the strongly correlated benefit evaluation index set, the multi-dimensional index classification of the multi-party soil erosion prevention, moisture conservation and fertilization full life cycle data stream is performed to obtain a multi-party index soil full life cycle data stream set; Based on the benefit evaluation model training framework set, the multi-indicator soil life cycle data stream set is respectively subjected to distributed training fusion to construct the global soil life cycle benefit evaluation model set.

3. The soil erosion prevention, moisture conservation and fertilization full life cycle benefit evaluation and optimization system according to claim 2, characterized in that: The obtaining of a set of strongly correlated benefit evaluation indicators includes: Sequentially extract any one evaluation index from the soil benefit evaluation index set; Based on any one of the evaluation indicators, a set of correlation coefficients of benefit indicators is obtained by traversing and calculating in the multi-party soil erosion prevention, moisture conservation and fertilization full life cycle data stream; The benefit index correlation coefficient set is weighted and integrated according to the soil benefit evaluation index set to obtain a benefit index correlation degree set; A preset correlation threshold is set, and the benefit indicator correlation set is screened and matched based on the preset correlation threshold to obtain the strongly correlated benefit evaluation indicator set.

4. The soil erosion prevention, moisture conservation and fertilization full life cycle benefit evaluation and optimization system as claimed in claim 3, characterized in that: The construction of the global soil life cycle benefit assessment model set includes: Based on the benefit evaluation model training framework set, the multi-indicator soil full life cycle data stream set is respectively distributedly trained to obtain a multi-framework multi-indicator benefit evaluation model set; Assigning weights to the set of strongly correlated benefit evaluation indicators according to the benefit indicator correlation degree set to determine a set of benefit indicator weight factors; Based on the benefit indicator weight factor set, the multi-framework multi-party indicator benefit evaluation model set is subjected to decision fusion to obtain a multi-framework multi-party benefit fusion model set; According to the benefit evaluation model training framework set, the model parameters of the multi-framework multi-party benefit fusion model set are extracted and aggregated for training, so as to construct the global soil full life cycle benefit evaluation model set.

5. The soil erosion prevention, moisture conservation and fertilization full life cycle benefit evaluation and optimization system according to claim 1, characterized in that: The key factors affecting the whole life cycle of soil are obtained, including: Acquire the soil erosion prevention, moisture conservation and fertilization data stream to be evaluated in the target area; Based on the local soil full life cycle benefit evaluation model, the soil erosion prevention, moisture conservation and fertilization data stream to be evaluated is evaluated, and soil full life cycle benefit evaluation parameters are output; Analyzing the influencing factors of the soil benefit evaluation index set to obtain a benefit index influencing factor set; Based on the soil benefit evaluation index set and the benefit index influencing factor set, a multi-objective benefit decision model is constructed; Based on the multi-objective benefit decision-making model, a key factor analysis is performed on the soil life cycle benefit evaluation parameters to obtain a set of key influencing factors of the soil life cycle.

6. The soil erosion prevention, moisture conservation and fertilization full life cycle benefit evaluation and optimization system according to claim 1, characterized in that: The soil benefit optimization management and control through the set of key influencing factors of the soil throughout its life cycle includes: Constructing a soil erosion prevention, moisture conservation and fertilization strategy library, matching the set of key influencing factors of the soil throughout its life cycle with the soil erosion prevention, moisture conservation and fertilization strategy library to obtain multiple soil optimization strategy solutions; Perform simulation evaluation based on the multiple soil optimization strategies to obtain benefit evaluation information of the multiple strategies, optimize the multiple soil optimization strategies based on the benefit evaluation information of the multiple strategies, and determine a target soil optimization strategy; Based on the target soil optimization strategy, soil efficiency optimization management and control is performed on the target area.

Images