Planting optimization method and system based on crop yield estimation in saline-alkali soil area

Abstract

The invention discloses a planting optimization method and system based on crop yield estimation in a saline-alkali land area, and relates to the field of crop yield prediction and planting optimization, and the method comprises the steps: data collection and fusion processing; performing dynamic prediction on soil water and salt migration, and realizing water and salt space-time distribution prediction by adopting a space-time cubic modeling self-adaptive model combined with physical correction; crop salt tolerance modeling: establishing a multi-interval salt tolerance model based on physiological response fitting and threshold segmentation, and introducing covariable correction; carrying out migration salt tolerance coupling modeling, introducing a salinity memory factor and a stage sensitivity regulation factor, and dynamically simulating growth response of crops under salinity stress; and crop yield estimation and planting optimization: generating a regional yield prediction and sowing configuration scheme based on coupling model output, and forming an optimization strategy in combination with measures such as irrigation and fertilization. According to the method, model-driven integrated optimization from yield estimation to planting decision of crops in the saline-alkali region is realized, and the method has high precision and wide applicability.

Description

Technical Field

[0001] This invention relates to the field of crop yield prediction and planting optimization technology, specifically to a planting optimization method and system based on crop yield prediction in saline-alkali areas. Background Technology

[0002] The planting optimization method and system based on crop yield prediction in saline-alkali areas is an agricultural information technology solution for saline-alkali farmland. By integrating remote sensing, monitoring, and historical data, and combining water and salt transport dynamic prediction with crop salt tolerance modeling, it achieves refined prediction of crop growth adaptability and yield levels in saline-alkali environments, and further generates scientifically reasonable planting optimization plans. Its role is to overcome the traditional extensive decision-making based on experience and single indicators, realizing intelligent optimization of the entire process from regional-scale yield prediction to field management measures, thereby improving the scientific nature, adaptability, and overall output efficiency of agricultural production in saline-alkali areas.

[0003] However, existing planting optimization methods based on crop yield prediction in saline-alkali areas suffer from technical problems such as a disconnect between yield prediction and planting optimization, and overly empirical optimization strategies. Existing methods for predicting soil water and salt transport dynamics suffer from technical problems such as a lack of physical constraints in the models, insufficient prediction accuracy and interpretability. Existing methods for modeling crop salt tolerance suffer from technical problems such as coarse threshold division and difficulty in reflecting the nonlinear response of crops under different salinity conditions. Existing coupled modeling methods for transport and salt tolerance suffer from technical problems such as only performing static coupling and not considering the temporal cumulative effect and stage sensitivity of crop growth. Summary of the Invention

[0004] To address the above issues and overcome the shortcomings of existing technologies, this invention provides a planting optimization method and system based on crop yield prediction in saline-alkali land areas. Addressing the technical problems of existing planting optimization methods based on crop yield prediction in saline-alkali land areas, such as a disconnect between yield prediction and planting optimization, and overly empirical optimization strategies, this solution creatively adopts a comprehensive progressive modeling and prediction method that combines soil water and salt transport dynamic prediction, crop salt tolerance modeling, and coupled transport and salt tolerance modeling. Furthermore, it uses yield prediction data to drive the generation of a comprehensive planting optimization strategy, thereby realizing a shift from "passive experience-based configuration" to "active optimization based on model deduction." "This transformation effectively improves the adaptability of crop planting and overall yield in saline-alkali areas. Addressing the technical problems of existing soil water and salt transport dynamic prediction methods, such as lack of physical constraints, insufficient prediction accuracy, and inadequate interpretability, this solution creatively adopts a spatiotemporal cubic modeling adaptive prediction model combined with physical correction. It extracts the spatiotemporal features of water and salt through a combination of 3D convolution and convolutional LSTM with deconvolution decoding. Simultaneously, it introduces a physical consistency loss function based on the water and salt diffusion equation for error constraint, achieving high-precision prediction that balances data-driven approaches and physical mechanisms, ensuring the long-term temporal and spatial distribution accuracy of water and salt transport prediction results." Physical rationality: Addressing the technical problems of existing crop salt tolerance modeling methods, such as coarse threshold division and difficulty in reflecting the nonlinear response of crops under different salinity conditions, this scheme creatively adopts an improved salt tolerance response method that combines crop physiological response fitting with segmented salt stress threshold modeling. It automatically divides salinity intervals through variation point detection and uses an exponential decay function and covariate weighted correction factors within each interval to finely characterize crop yield responses under different salinities. This achieves a shift in salt tolerance modeling from a coarse description using a single threshold to a multi-interval, adjustable, fine-grained response model, improving the model's adaptability to different crops and ecological environments. To address the technical problem in existing coupled modeling methods for salt tolerance migration that only perform static coupling and do not consider the temporal cumulative effects and stage sensitivity during crop growth, this solution creatively adopts a temporal coupled modeling method that combines dynamic water and salt simulation with crop response regulation mechanisms. It introduces an exponentially decaying salt memory factor and a stage sensitivity regulation factor, enabling the model to dynamically characterize the accumulation and recovery process of salt stress on a temporal scale. Furthermore, it is dynamically invoked in conjunction with the crop salt tolerance response model, ultimately achieving dynamic response simulation of the entire crop growth process in saline-alkali areas, thereby significantly improving the accuracy and reliability of yield prediction.

[0005] The technical solution adopted by this invention is as follows: The planting optimization method based on crop yield prediction in saline-alkali areas provided by this invention includes the following steps:

[0006] Step S1: Data acquisition, fusion, and processing;

[0007] Step S2: Prediction of soil water and salt transport dynamics;

[0008] Step S3: Crop salt tolerance modeling;

[0009] Step S4: Modeling the coupling of transport and salt tolerance;

[0010] Step S5: Crop yield forecasting and planting optimization.

[0011] Furthermore, in step S1, the data acquisition and fusion processing is used to collect multi-source saline-alkali land soil and meteorological data and integrate the original dataset. Specifically, it is to obtain a standardized fusion dataset through multi-source data acquisition and standardization and processing procedures.

[0012] The multi-source data acquisition includes remote sensing data acquisition, field monitoring data acquisition, and retrieval of historical statistical data.

[0013] Further, in step S2, the soil water and salt transport dynamic prediction is used to construct a spatiotemporal prediction model to simulate the dynamic distribution process of salt and water in the soil. Specifically, based on the standardized fusion dataset, an adaptive prediction model combining physical correction and spatiotemporal cubic modeling is used to predict the soil water and salt transport dynamics, obtaining water and salt transport dynamic data, including the following steps:

[0014] Step S21: Data resampling, specifically, by unifying the spatial resolution and time step, interpolating, repairing missing values ​​and rasterizing the original soil water and salt data, and reconstructing it through spatiotemporal standardization to obtain a standardized tensor sequence;

[0015] Step S22: Spatiotemporal cube structure modeling, specifically by encapsulating the spatial raster tensor of continuous time steps into spatiotemporal cube input units, constructing a cube data structure, and obtaining the spatiotemporal input tensor;

[0016] Step S23: Spatiotemporal feature extraction and encoding, specifically, performing convolution calculations on the spatiotemporal cube through a three-dimensional convolutional network to encode spatiotemporal features and obtain latent spatial state features;

[0017] Step S24: Deconvolutional Long Short-Term Memory Network Decoding, specifically by inputting the latent spatial state features into the convolutional long short-term memory network and combining it with the deconvolutional layer for decoding, performing temporal prediction and recovery, and obtaining the water and salt state prediction frame for the future time step;

[0018] Step S25: Construct a physical constraint loss function. Specifically, by introducing the water-salt diffusion equation and conservation relationship, a regularization constraint based on mass conservation and diffusion gradient is constructed to correct the prediction error and obtain the physical consistency loss.

[0019] Step S26: Prediction of soil water and salt transport dynamics, specifically by jointly minimizing the mean square error loss and physical consistency loss, to train and infer the network model and obtain water and salt transport dynamic data.

[0020] Further, in step S3, the crop salt tolerance modeling is used to establish growth adaptation models for different crops under different salinity conditions based on crop varieties and growth data. Specifically, based on the standardized fusion dataset, an improved salt tolerance response method combining crop physiological response fitting and salt stress threshold segmentation modeling is used to model crop salt tolerance, resulting in a crop salt tolerance response model, including the following steps:

[0021] Step S31: Input factor extraction, specifically, by extracting the target crop yield value as the response variable, and extracting soil salinity, water content, air temperature, crop type code and sowing time to form an input matrix, constructing variables and factors, and obtaining the original input factor matrix and observed response variables;

[0022] Step S32: Salt stress threshold interval division, specifically, the soil salinity level is divided into multiple threshold segments by using the mutation point detection algorithm, and the salinity interval is segmented to obtain a set of salinity thresholds;

[0023] Step S33: Physiological response factor function modeling, specifically, constructing a nonlinear response function based on exponential decay and baseline correction within each salinity interval, performing interval fitting calculations, and obtaining the piecewise salinity response function;

[0024] Step S34: Covariate coupling response correction, specifically by constructing weighted modulation factors of moisture content, air temperature, sowing time and crop type, and correcting the salinity response function to obtain the covariate coupling correction model;

[0025] Step S35: Parameter optimization model training, specifically by minimizing the yield prediction error loss function and combining regularization constraints to perform parameter optimization training, and obtain the final crop salt tolerance response model.

[0026] Further, in step S4, the transport salt tolerance coupling modeling is used to integrate soil water and salt changes with the crop salt tolerance model to simulate the response characteristics of crops in a dynamic salt environment. Specifically, based on the water and salt transport dynamic data and the crop salt tolerance response model, a time-series coupling modeling method combining dynamic water and salt simulation and crop response regulation mechanism is used to perform transport salt tolerance coupling modeling to obtain a coupled prediction model, including the following steps:

[0027] Step S41: Construction of dynamic water and salt state sequence, specifically by extracting salt and water data during the crop growth period, constructing time-step driving variables, and obtaining water and salt driving matrix;

[0028] Step S42: Design of physiological response memory factors, specifically by constructing a salt stress accumulation index and using an exponential decay weighting function to model the crop salt stress memory effect and obtain time step memory factors;

[0029] Step S43: Design of stage-sensitive regulatory factors, specifically by identifying different growth stages of crops and setting stage sensitivity coefficients, and then modifying the salt sensitivity to obtain stage-sensitive regulatory factors.

[0030] Step S44: Dynamic coupling of crop response model, specifically by calling the crop salt tolerance response model, combining time step memory factor and stage regulation factor, to perform dynamic yield calculation and obtain dynamic response yield sequence;

[0031] Step S45: Coupled prediction model construction, specifically, by calibrating and optimizing the dynamic response yield sequence based on the principle of minimizing yield prediction error, to obtain the final coupled prediction model.

[0032] Further, in step S5, the crop yield prediction and planting optimization are used to calculate the expected yield of crops planted in a specific area based on the output of the coupled model and environmental parameters. Specifically, the coupled prediction model is used to predict crop yield to obtain regional yield prediction data. By analyzing the spatial distribution of the regional yield prediction data, high-yield and low-yield difference areas are identified. Combined with the salt tolerance level and yield threshold of different crop varieties, variety selection and matching are performed. Then, based on the spatial distribution characteristics and variety matching results, a sowing area configuration plan is generated. Combined with water and salt transport dynamic data, field management measures such as irrigation, fertilization and salt removal are scheduled and optimized to comprehensively output regional planting optimization strategies and obtain planting optimization data.

[0033] The planting optimization system based on crop yield prediction in saline-alkali areas provided by the present invention includes a data acquisition and fusion processing module, a soil water and salt transport dynamic prediction module, a crop salt tolerance modeling module, a transport and salt tolerance coupled modeling module, and a planting optimization module.

[0034] The data acquisition and fusion processing module is used for data acquisition and fusion processing. Through data acquisition and fusion processing, a standardized fusion dataset is obtained, and the standardized fusion dataset is sent to the soil water and salt transport dynamic prediction module and the crop salt tolerance modeling module.

[0035] The soil water and salt transport dynamic prediction module is used to predict soil water and salt transport dynamics. Through the prediction of soil water and salt transport dynamics, water and salt transport dynamic data is obtained, and the water and salt transport dynamic data is sent to the transport salt tolerance coupling modeling module.

[0036] The crop salt tolerance modeling module is used for crop salt tolerance modeling. Through crop salt tolerance modeling, a crop salt tolerance response model is obtained, and the crop salt tolerance response model is sent to the transport salt tolerance coupling modeling module.

[0037] The transport salt tolerance coupling modeling module is used for transport salt tolerance coupling modeling. Through transport salt tolerance coupling modeling, a coupling prediction model is obtained, and the coupling prediction model is sent to the planting optimization module.

[0038] The planting optimization module is used for crop yield prediction and planting optimization. Through crop yield prediction and planting optimization, regional yield prediction data and planting optimization strategies are obtained.

[0039] The beneficial effects achieved by the present invention using the above solution are as follows:

[0040] (1) In response to the technical problems of existing crop yield prediction optimization methods based on saline-alkali areas, such as the disconnect between yield prediction and planting optimization and the overly empirical optimization strategies, for example, in traditional saline-alkali area improvement planting, the calculation is usually based solely on soil salinity index or single crop trial yield, lacking comprehensive consideration of the dynamic changes in salinity and water in different regions and the differences in crop salt tolerance, resulting in the regional scale sowing structure and field management plan often lacking scientificity and universality; This scheme creatively adopts a comprehensive progressive modeling prediction method that combines soil water and salt transport dynamic prediction, crop salt tolerance modeling and transport salt tolerance coupling modeling, and drives the generation of comprehensive planting optimization strategies through yield prediction data, thereby realizing the transformation from "passive experience configuration" to "active optimization based on model deduction", effectively improving the adaptability of crop planting and the overall yield level in saline-alkali areas;

[0041] (2) In view of the technical problems in existing soil water and salt transport dynamic prediction methods, such as the lack of physical constraints, insufficient prediction accuracy and interpretability, prediction methods based on conventional machine learning or simple regression models can only fit the static correlation between historical salt content and rainfall and groundwater depth, and cannot guarantee that the prediction results meet the salt conservation or water diffusion law. Therefore, the deviation will gradually accumulate in the long-term prediction and is difficult to use for agricultural decision-making. This scheme creatively adopts a spatiotemporal cubic modeling adaptive prediction model with physical correction. It extracts the spatiotemporal features of water and salt by combining three-dimensional convolution and convolutional LSTM with deconvolution decoding. At the same time, it introduces a physical consistency loss function based on the water and salt diffusion equation for error constraint, realizing high-precision prediction that takes into account both data-driven and physical mechanisms, and ensuring the physical rationality of water and salt transport prediction results in long-term temporal and spatial distribution.

[0042] (3) In response to the technical problems in existing crop salt tolerance modeling methods, such as coarse threshold division and difficulty in reflecting the nonlinear response of crops under different salinity conditions, traditional studies usually use a single "salinity critical value" to classify whether crops are salt-tolerant, while ignoring the typical segmented characteristics that yield is almost unaffected in the low-salinity range, yield decreases slowly with salinity in the medium-salinity range, and yield decreases sharply in the high-salinity range, which makes the model unable to accurately characterize the yield-salinity relationship. This scheme creatively adopts an improved salt tolerance response method that combines crop physiological response fitting with salt stress threshold segmentation modeling. It automatically divides the salinity range through the detection of variation points, and uses an exponential decay function and covariate weighted correction factor within the range to finely characterize the crop yield response under different salinities. This realizes the transformation of salt tolerance modeling from a coarse description with a single threshold to a fine response modeling with multiple ranges and adjustable parameters, and improves the adaptability of the model to different crops and different ecological environments.

[0043] (4) In the existing coupled modeling methods for transport salt tolerance, there are technical problems such as only static coupling and failure to consider the time-series cumulative effect and stage sensitivity of crop growth. For example, some studies only combine average salinity with yield regression, but fail to reflect the differences in salt stress at different stages such as seedling, tillering and grain filling stages, and do not introduce the historical memory effect of salt stress. Therefore, the predicted results often differ greatly from the measured yield. This scheme creatively adopts a time-series coupled modeling method that combines dynamic water and salt simulation with crop response regulation mechanism, and introduces exponentially decaying salt memory factor and stage sensitivity regulation factor, so that the model can dynamically depict the accumulation and recovery process of salt stress on a time scale, and dynamically call it in combination with crop salt tolerance response model, and finally realize the dynamic response simulation of the entire growth process of crops in saline-alkali areas, thereby significantly improving the accuracy and reliability of yield prediction. Attached Figure Description

[0044] Figure 1 This is a flowchart illustrating the planting optimization method based on crop yield prediction in saline-alkali areas provided by the present invention.

[0045] Figure 2 A schematic diagram of the planting optimization system based on crop yield prediction in saline-alkali areas provided by the present invention;

[0046] Figure 3 This is a flowchart illustrating the process of predicting the dynamics of soil water and salt transport in step S2.

[0047] Figure 4 A schematic diagram of the process for modeling crop salt tolerance in step S3;

[0048] Figure 5 A schematic diagram of the process for modeling the salt tolerance coupling of step S4.

[0049] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the embodiments of the invention to explain the invention and do not constitute a limitation thereof. Detailed Implementation

[0050] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0051] In the description of this invention, it should be understood that the terms "upper", "lower", "front", "rear", "left", "right", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0052] Example 1, see Figure 1 The present invention provides a planting optimization method based on crop yield prediction in saline-alkali areas, which includes the following steps:

[0053] Step S1: Data acquisition, fusion, and processing;

[0054] Step S2: Prediction of soil water and salt transport dynamics;

[0055] Step S3: Crop salt tolerance modeling;

[0056] Step S4: Modeling the coupling of transport and salt tolerance;

[0057] Step S5: Crop yield forecasting and planting optimization.

[0058] By performing the above operations, this approach addresses the technical problems in existing crop yield prediction optimization methods for saline-alkali areas, such as the disconnect between yield prediction and planting optimization, and the overly empirical nature of optimization strategies. For example, in traditional saline-alkali area improvement planting, calculations are usually based solely on soil salinity indicators or the yield of a single crop trial, lacking comprehensive consideration of the dynamic changes in salinity and water content and the differences in crop salt tolerance in different regions. This results in regional-scale sowing structures and field management plans often lacking scientific rigor and universality. This solution creatively adopts a comprehensive progressive modeling and prediction method that combines soil water and salt transport dynamic prediction, crop salt tolerance modeling, and transport-salt tolerance coupled modeling. It also drives the generation of comprehensive planting optimization strategies through yield prediction data, thereby realizing a shift from "passive experience-based configuration" to "active optimization based on model deduction," effectively improving the adaptability of crop planting and the overall yield level in saline-alkali areas.

[0059] Example 2, see Figure 1 and Figure 2 This embodiment is based on the above embodiment. In step S1, the data acquisition and fusion processing is used to collect multi-source saline-alkali soil and meteorological data and integrate the original dataset. Specifically, it is to obtain a standardized fusion dataset through multi-source data acquisition and standardization and processing procedures.

[0060] The multi-source data acquisition includes remote sensing data acquisition, field monitoring data acquisition, and retrieval of historical statistical data.

[0061] Preferably, the remote sensing data acquisition mainly includes surface reflectance, vegetation index (NDVI), surface temperature, and evapotranspiration; the field monitoring data acquisition mainly includes soil electrical conductivity, pH value, salinity, moisture content, organic matter content, soil type, groundwater depth, soil physicochemical characteristic parameters, and real-time meteorological data, including rainfall, wind speed, evaporation, and temperature and humidity.

[0062] The historical statistical data includes recent regional crop planting conditions, historical weather change trends, and historical yield data.

[0063] Example 3, see Figure 1 , Figure 2 and Figure 3 This embodiment is based on the above embodiment. In step S2, the soil water and salt transport dynamic prediction is used to construct a spatiotemporal prediction model to simulate the dynamic distribution process of salt and water in the soil. Specifically, based on the standardized fusion dataset, an adaptive prediction model combining physical correction and spatiotemporal cubic modeling is used to predict the soil water and salt transport dynamics and obtain water and salt transport dynamic data. This includes the following steps:

[0064] Step S21: Data resampling, specifically, by unifying the spatial resolution and time step, interpolating, repairing missing values ​​and rasterizing the original soil water and salt data, and reconstructing it through spatiotemporal standardization to obtain a standardized tensor sequence;

[0065] Preferably, the unified spatial resolution is specifically achieved by using bilinear interpolation or kriging interpolation to unify soil salinity and moisture observations with different spatial resolutions to a standard grid of 30m×30m.

[0066] The time step is preferably 1 day or 1 week, and the missing value repair is performed by linear interpolation and sliding window mean to repair the missing time points.

[0067] The standardized tensor sequence specifically refers to water and salt raster data that has undergone unified spatial gridding;

[0068] Step S22: Spatiotemporal cube structure modeling, specifically by encapsulating the normalized tensor sequence of continuous time steps into spatiotemporal cube input units, constructing a cube data structure, and obtaining the spatiotemporal input tensor;

[0069] Preferably, the spatiotemporal cube input unit is a data structure that unifies the time dimension, spatial dimension, and feature dimension; the spatiotemporal input tensor encapsulates soil salinity distribution, moisture distribution, and related meteorological and soil characteristic data at multiple consecutive time points within a preset time window; spatially, it is organized in the form of a regular grid, with each grid unit corresponding to a specific geographical location; in terms of features, each grid unit contains multiple observation indicators, including soil salinity concentration, water content, electrical conductivity, temperature, and evapotranspiration; temporally, the spatiotemporal cube retains the dynamic changes of the above observation indicators at multiple consecutive time points.

[0070] Through the above encapsulation, the spatiotemporal cube can not only represent the spatial distribution state at a certain moment, but also reflect the evolution trend and dynamic change relationship between different time points, thereby realizing the coupled modeling of time series information and spatial grid information, ensuring that the subsequent three-dimensional convolution calculation can simultaneously capture temporal dependent features and spatially related features, and providing complete spatiotemporal input information for the prediction of soil water and salt transport.

[0071] Step S23: Spatiotemporal feature extraction and encoding, specifically, performing convolution calculations on the spatiotemporal cube through a three-dimensional convolutional network to encode spatiotemporal features and obtain latent spatial state features;

[0072] Preferably, the structural parameters of the three-dimensional convolutional network are set as follows: the kernel size is 3×3×3, the stride is preferably 1, and the padding method is Same padding to ensure that the spatial size of the output tensor after convolution is consistent with that of the input; the number of convolutional layers is set to 3, and the number of output channels is 64, 128 and 128 respectively; normalization processing and nonlinear activation functions can be selected between each layer to enhance the stability of feature extraction and nonlinear expression ability.

[0073] Step S24: Deconvolutional Long Short-Term Memory Network Decoding, specifically by inputting the latent spatial state features into the convolutional long short-term memory network and combining it with the deconvolutional layer for decoding, performing temporal prediction and recovery, and obtaining the water and salt state prediction frame for the future time step;

[0074] The calculation formula for the time-series prediction recovery is as follows:

[0075] ;

[0076] In the formula, It is the predicted water-salt state tensor sequence for the next h time steps, used to represent the predicted water-salt state frames for those future time steps. Deconv is a deconvolutional layer operation, ConvLSTM is a convolutional long short-term memory network operation, and f 3D-CNN It is a 3D convolutional network operation, C t It is a spatiotemporal input tensor, where t is the time index and h is the prediction time step index;

[0077] Preferably, the structural parameters of the convolutional long short-term memory network are set as follows: the number of network layers is set to 2, the number of hidden units in each layer is 128, the kernel size is preferably 3×3, and the padding method is Same padding to maintain the consistency of temporal features in the convolution operation; at the decoding end, a deconvolution operation is used with a kernel size of 3×3 and a stride of 2 to restore the spatial resolution after downsampling to the original input size, thereby obtaining the water and salt state prediction frame for the future time step;

[0078] Step S25: Construct a physical constraint loss function. Specifically, by introducing the water-salt diffusion equation and conservation relationship, a regularization constraint based on mass conservation and diffusion gradient is constructed to correct the prediction error and obtain the physical consistency loss.

[0079] The formula for calculating the physical consistency loss is as follows:

[0080] ;

[0081] In the formula, This is the physical consistency loss, where t is the time index. It is a predicted salinity field used to represent predicted salt concentration values. It is the spatial gradient operator, D s The effective diffusion coefficient of salt is preferably taken from 0.01 to 0.1 m² / day. It is obtained through indoor non-convective diffusion experiments. Specifically, under constant hydraulic conditions, the diffusion profile of salt concentration in soil over time and space is measured, and the effective diffusion coefficient is obtained by fitting the Fick diffusion equation. It predicts the water content field, D w The hydraulic dispersion coefficient is preferably within the range of 0.1 to 1.0 m² / day. It is obtained through a constant flow breakthrough test, specifically by injecting an inert tracer under constant flow conditions, collecting the outlet breakthrough curve, fitting the convection-dispersion equation to invert the dispersion parameter, and then calculating it. It is the L2 norm square operator;

[0082] Step S26: Prediction of soil water and salt transport dynamics, specifically by jointly minimizing the mean square error loss and physical consistency loss, to train and infer the network model and obtain water and salt transport dynamic data;

[0083] Preferably, the formula for calculating the joint minimization of mean square error loss and physical consistency loss is as follows:

[0084] ;

[0085] In the formula, L is the joint loss function, L MSE It is to minimize the mean squared error loss. This is the weight for physical consistency loss, with a preferred value range of 0.1 to 0.5;

[0086] Preferably, Table 1 is an example table of network model training parameters. As shown in the table, the spatial grid resolution with uniform spatial resolution is 30m × 30m to balance the accessibility of remote sensing data with the accuracy of field-level management; the time window length is set to 8 time steps, which are stacked continuously to form a cubic input unit, covering the water and salt changes of a typical agricultural cycle; the prediction step length is set to 14 days to connect with management cycles such as irrigation and salt drainage, facilitating yield prediction and field scheduling.

[0087] In terms of network structure, the 3D convolution uses a 3 × 3 × 3 convolution kernel to capture local spatiotemporal features. The number of convolution channels is 64, 128 and 128 respectively. The number of long short-term memory network units in the convolution is set to 128 to fully model temporal dependencies.

[0088] The training process uses the Adam optimizer with an initial learning rate of 3e-4 and a batch size of 16 to balance parameter update efficiency and memory consumption. A physical consistency constraint is introduced into the loss function, with a weight coefficient α of 0.3, to ensure that the prediction results conform to the water and salt conservation law. The training epochs are capped at 60, supplemented by an early stopping mechanism with a patience value of 8. Training is terminated early when the validation set error no longer decreases within a certain number of consecutive epochs, thus ensuring stable model convergence while avoiding overfitting.

[0089] Table 1. Example of network model training parameters

[0090] By performing the above operations, this approach addresses the technical problems of existing soil water and salt transport dynamic prediction methods, such as the lack of physical constraints in the models and insufficient prediction accuracy and interpretability. For example, prediction methods based on conventional machine learning or simple regression models can often only fit the static correlation between historical salinity content and rainfall and groundwater depth, but cannot guarantee that the prediction results meet the laws of salt conservation or water diffusion. Therefore, the bias will gradually accumulate in the long-term prediction, making it difficult to use for agricultural decision-making. This solution creatively adopts a spatiotemporal cubic modeling adaptive prediction model that combines physical correction. It extracts the spatiotemporal features of water and salt by combining three-dimensional convolution and convolutional LSTM with deconvolution decoding. At the same time, it introduces a physical consistency loss function based on the water and salt diffusion equation for error constraint, achieving high-precision prediction that takes into account both data-driven and physical mechanisms, and ensuring the physical rationality of water and salt transport prediction results in long-term temporal and spatial distribution.

[0091] Example 4, see Figure 1 , Figure 2 and Figure 4 This embodiment is based on the above embodiment. In step S3, the crop salt tolerance modeling is used to establish growth adaptation models of different crops under different salinity conditions based on crop varieties and growth data. Specifically, based on the standardized fusion dataset, a salt tolerance response improvement method combining crop physiological response fitting and salt stress threshold segmentation modeling is used to perform crop salt tolerance modeling to obtain a crop salt tolerance response model, including the following steps:

[0092] Step S31: Input factor extraction, specifically, by extracting the target crop yield value as the response variable, and extracting soil salinity, water content, air temperature, crop type code and sowing time to form an input matrix, constructing variables and factors, and obtaining the original input factor matrix and observed response variables;

[0093] Step S32: Salt stress threshold interval division, specifically, the soil salinity level is divided into multiple threshold segments by using the mutation point detection algorithm, and the salinity interval is segmented to obtain a set of salinity thresholds;

[0094] Preferably, the mutation point detection algorithm is specifically implemented based on a combination of Bayesian information criterion and dynamic programming. It is used to automatically identify significant inflection points on the curve of crop yield change with salinity, thereby determining typical threshold intervals such as low salinity, medium salinity and high salinity. Preferably, the number of segments is 3 to 5 to ensure that the segments are both biologically meaningful and avoid overfitting.

[0095] Step S33: Physiological response factor function modeling, specifically, constructing a nonlinear response function based on exponential decay and baseline correction within each salinity interval, performing interval fitting calculations, and obtaining the piecewise salinity response function;

[0096] The formula for calculating the segmented salinity response function is as follows:

[0097] ;

[0098] In the formula, f(·) is the piecewise salinity response function, S is the salinity index used to characterize salt concentration, k is the number of salinity segments, and i is the salinity segment index. It is an indicator function. It is a set of salinity thresholds, where, f is the salinity threshold corresponding to the i-th salinity segment. i (S) is the physiological response function;

[0099] Preferably, the formula for calculating the physiological response function is:

[0100] ;

[0101] In the formula, f i (S) is the physiological response function, a i It is the initial yield level of the interval, preferably the average observed yield at the starting point of the salinity interval, b i This is the sensitivity coefficient of yield to changes in salinity, preferably ranging from 0.01 to 0.5, used to adjust the rate of decrease in the index. S is a salinity index used to characterize salt concentration. i It is a baseline correction term, preferably greater than zero, to ensure that when salinity continues to rise, the predicted yield will not approach a negative value, but will stabilize at a certain minimum physiological yield level.

[0102] Step S34: Covariate coupling response correction, specifically by constructing weighted modulation factors of moisture content, air temperature, sowing time and crop type, and correcting the salinity response function to obtain the covariate coupling correction model;

[0103] The formula for calculating the weighted modulation factor is as follows:

[0104] ;

[0105] In the formula, These are weighted modulation factors. w1 is the moisture content weight, with a default value of 0.3. W is the moisture content factor. w2 is the temperature weight, with a default value of 0.3. T is the temperature factor. w3 is the sowing time weight, with a default value of 0.2. Ts is the sowing time factor. w4 is the crop type weight, with a default value of 0.2. C is the crop type factor.

[0106] The calculation formula for the covariate coupling correction model is as follows:

[0107] ;

[0108] In the formula, Y pred This is the predicted output of the covariate coupling modified model, where f(·) is the piecewise salinity response function. This is the coupling strength coefficient, preferably ranging from 0.1 to 0.5. It is a weighted modulation factor;

[0109] Step S35: Parameter optimization model training, specifically by minimizing the yield prediction error loss function and combining regularization constraints to perform parameter optimization training, and obtain the final crop salt tolerance response model;

[0110] Preferably, the loss function for minimizing the output prediction error is the mean squared error (MSE), the regularization constraint is L2 regularization to prevent overfitting, the optimizer is the Adam optimizer, the learning rate is preferably 1e-3, the training batch size is preferably 16 to 32, the number of training epochs is preferably 50 to 100, and an early stopping strategy is introduced, which terminates training early when the validation set error no longer decreases within a certain number of consecutive epochs, thereby improving the generalization performance of the model.

[0111] By performing the above operations, this approach addresses the technical problems in existing crop salt tolerance modeling methods, such as coarse threshold division and difficulty in reflecting the nonlinear response of crops under different salinity conditions. For example, traditional studies often use a single "salinity threshold" to classify whether a crop is salt-tolerant, ignoring the typical segmented characteristics of yield being almost unaffected in low-salinity ranges, yield decreasing slowly with salinity in medium-salinity ranges, and yield decreasing sharply in high-salinity ranges. This results in the model being unable to accurately characterize the yield-salinity relationship. This solution creatively adopts an improved salt tolerance response method that combines crop physiological response fitting with segmented salt stress threshold modeling. It automatically divides salinity ranges through mutation point detection and uses an exponential decay function and covariate weighted correction factor within each range to finely characterize the crop yield response under different salinities. This achieves a shift in salt tolerance modeling from a coarse description using a single threshold to a fine, adjustable multi-range response modeling, improving the model's adaptability to different crops and different ecological environments.

[0112] Example 5, see Figure 1, Figure 2 and Figure 5 This embodiment is based on the above embodiment. In step S4, the transport salt tolerance coupling modeling is used to integrate soil water and salt changes with crop salt tolerance models to simulate the response characteristics of crops in a dynamic salt environment. Specifically, based on the water and salt transport dynamic data and the crop salt tolerance response model, a time-series coupling modeling method combining dynamic water and salt simulation and crop response regulation mechanisms is used to perform transport salt tolerance coupling modeling to obtain a coupled prediction model, including the following steps:

[0113] Step S41: Construction of dynamic water and salt state sequence, specifically by extracting salt and water data during the crop growth period, constructing time-step driving variables, and obtaining water and salt driving matrix;

[0114] Preferably, the water-salt driving matrix adopts a 30 m × 30 m grid in space, consistent with step S21, and adopts a time window consistent with the crop growth period in time, with a daily or weekly scale as the time step; wherein the driving variables of each time step include root zone soil salt concentration (expressed as electrical conductivity EC), water content and rainfall evapotranspiration factor.

[0115] Step S42: Design of physiological response memory factors, specifically by constructing a salt stress accumulation index and using an exponential decay weighting function to model the crop salt stress memory effect and obtain time step memory factors;

[0116] The formula for calculating the time step memory factor is as follows:

[0117] ;

[0118] In the formula, R t This is the time step memory factor, used to represent the delayed effect of past high-salinity periods on current growth, where t is the time index. It is a time-step memory index. It is a time step Time step weights It is a time step The root region salinity is specifically represented by the conductivity EC, and f(·) is the piecewise salinity response function;

[0119] The formula for calculating the time step weight is as follows:

[0120] ;

[0121] In the formula, It is the attenuation coefficient, preferably ranging from 0.01 to 0.1, used to control the rate at which historical salinity stress decays over time;

[0122] Step S43: Design of stage-sensitive regulatory factors, specifically by identifying different growth stages of crops and setting stage sensitivity coefficients, and then modifying the salt sensitivity to obtain stage-sensitive regulatory factors.

[0123] Preferably, the growth stages are divided into segments according to agronomic methods, specifically into seedling stage, heading stage, and maturity stage; the preferred range of sensitivity coefficients for each stage is: 0.6 to 0.8 for seedling stage, 0.4 to 0.6 for heading stage, and 0.2 to 0.3 for maturity stage, to reflect the differentiated sensitivity of different growth stages to salt stress;

[0124] Step S44: Dynamic coupling of crop response model, specifically by calling the crop salt tolerance response model, combining time step memory factor and stage regulation factor, to perform dynamic yield calculation and obtain dynamic response yield sequence;

[0125] Preferably, the formula for calculating the dynamic output is:

[0126] ;

[0127] In the formula, Y t It is the predicted output contribution value. This is the output of the crop salt tolerance response model under the current salinity conditions, S t It is the salinity index at time t, used to represent the current salinity conditions. This is the memory effect weighting coefficient, preferably ranging from 0.1 to 0.5, R. t It is a time-step memory factor. It is a stage-sensitive regulatory factor

[0128] Step S45: Coupled prediction model construction, specifically, by calibrating and optimizing the dynamic response yield sequence based on the principle of minimizing yield prediction error, the final coupled prediction model is obtained;

[0129] Preferably, the yield prediction error is measured using root mean square error (RMSE) or mean absolute percentage error (MAPE), and the parameters are iteratively updated using the Adam optimizer, with a learning rate preferably 1e-3. At the same time, a regularization term is introduced, preferably L2 regularization, with a regularization coefficient ranging from 1e-4 to 1e-2, to prevent the model from overfitting.

[0130] By performing the above operations, this approach addresses the technical problems inherent in existing coupled modeling methods for salt tolerance migration. These methods often only perform static coupling and fail to consider the temporal cumulative effects and stage sensitivity of crop growth. For example, some studies only combine average salinity with yield regression, failing to reflect the differences in salt stress response at different stages such as seedling, tillering, and grain-filling stages, and also neglecting to incorporate the historical memory effect of salt stress. Consequently, the predicted results often differ significantly from the measured yield. This approach creatively employs a temporal coupled modeling method that combines dynamic water-salt simulation with crop response regulation mechanisms. It introduces an exponentially decaying salt memory factor and a stage sensitivity regulation factor, enabling the model to dynamically characterize the accumulation and recovery process of salt stress on a temporal scale. Furthermore, it dynamically incorporates crop salt tolerance response models, ultimately achieving dynamic response simulation of the entire crop growth process in saline-alkali areas. This significantly improves the accuracy and reliability of yield prediction.

[0131] Example 6, see Figure 1 and Figure 2 This embodiment is based on the above embodiment. In step S5, the crop yield prediction and planting optimization are used to calculate the expected yield of crops planted in a specific area based on the output of the coupled model and environmental parameters. Specifically, the coupled prediction model is used to predict crop yield to obtain regional yield prediction data. By analyzing the spatial distribution of the regional yield prediction data, high-yield and low-yield difference areas are identified. Combined with the salt tolerance level and yield threshold of different crop varieties, variety selection and matching are performed. Then, based on the spatial distribution characteristics and variety matching results, a sowing area configuration plan is generated. Combined with water and salt transport dynamic data, the field management measures of irrigation, fertilization and salt removal are scheduled and optimized, thereby comprehensively outputting regional planting optimization strategies and obtaining planting optimization data.

[0132] Preferably, the field management scheduling optimization is quantitatively modeled by constructing a multi-objective optimization function, with the main optimization objectives being to increase the total regional yield, reduce resource input costs, and maintain water and salt balance. The optimization function is then iterated under constraints, which specifically include: the total cultivated land area of ​​the region must be controlled within the planning range, the sowing ratio of various crops should meet the requirements of agricultural structure, the total amount of irrigation and fertilization should not exceed the available resource threshold, and the salinity level should be lower than the crop tolerance limit.

[0133] Under the premise of satisfying the above constraints, the generated sowing area configuration scheme is preferably solved iteratively using an improved multi-objective evolutionary algorithm, so as to obtain a field management scheme that takes into account yield, cost and risk.

[0134] Example 7, see Figure 1 and Figure 2This embodiment is based on the above embodiment. The planting optimization system based on crop yield prediction in saline-alkali areas provided by the present invention includes a data acquisition and fusion processing module, a soil water and salt transport dynamic prediction module, a crop salt tolerance modeling module, a transport and salt tolerance coupled modeling module, and a planting optimization module.

[0135] The data acquisition and fusion processing module is used for data acquisition and fusion processing. Through data acquisition and fusion processing, a standardized fusion dataset is obtained, and the standardized fusion dataset is sent to the soil water and salt transport dynamic prediction module and the crop salt tolerance modeling module.

[0136] The soil water and salt transport dynamic prediction module is used to predict soil water and salt transport dynamics. Through the prediction of soil water and salt transport dynamics, water and salt transport dynamic data is obtained, and the water and salt transport dynamic data is sent to the transport salt tolerance coupling modeling module.

[0137] The crop salt tolerance modeling module is used for crop salt tolerance modeling. Through crop salt tolerance modeling, a crop salt tolerance response model is obtained, and the crop salt tolerance response model is sent to the transport salt tolerance coupling modeling module.

[0138] The transport salt tolerance coupling modeling module is used for transport salt tolerance coupling modeling. Through transport salt tolerance coupling modeling, a coupling prediction model is obtained, and the coupling prediction model is sent to the planting optimization module.

[0139] The planting optimization module is used for crop yield prediction and planting optimization. Through crop yield prediction and planting optimization, regional yield prediction data and planting optimization strategies are obtained.

[0140] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0141] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention.

[0142] The present invention and its embodiments have been described above. This description is not restrictive, and the accompanying drawings are only one embodiment of the present invention; the actual structure is not limited thereto. In conclusion, if those skilled in the art are inspired by this description and design similar structures and embodiments without departing from the spirit of the invention, such designs should fall within the protection scope of the present invention.

Claims

1. A planting optimization method based on crop yield prediction in saline-alkali areas, characterized by: The method includes the following steps: Step S1: Data acquisition and fusion processing. Through multi-source data acquisition and standard processing procedures, a standardized fusion dataset is obtained. Step S2: Soil water and salt transport dynamics prediction. Based on the standardized fusion dataset, an adaptive prediction model combining physical correction and spatiotemporal cubic modeling is used to predict soil water and salt transport dynamics and obtain water and salt transport dynamic data. Step S3: Crop salt tolerance modeling. Based on the standardized fusion dataset, an improved salt tolerance response method combining crop physiological response fitting and salt stress threshold segmentation modeling is used to model crop salt tolerance, resulting in a crop salt tolerance response model. This includes the following steps: Step S31: Input factor extraction; Step S32: Salt stress threshold interval division; Step S33: Physiological response factor function modeling; Step S34: Covariate coupling response correction; Step S35: Parameter optimization model training. Step S4: Coupled modeling of salt tolerance migration. Based on the dynamic data of water and salt migration and the crop salt tolerance response model, a time-series coupled modeling method combining dynamic water and salt simulation and crop response regulation mechanism is used to perform coupled modeling of salt tolerance migration to obtain a coupled prediction model. Step S5: Crop yield prediction and planting optimization. Using the coupled prediction model, crop yield is predicted to obtain regional yield prediction data, and regional planting optimization strategies are output to obtain planting optimization data.

2. The planting optimization method based on crop yield prediction in saline-alkali areas according to claim 1, characterized in that: In step S1, the multi-source data acquisition includes remote sensing data acquisition, field monitoring data acquisition, and retrieval of historical statistical data.

3. The planting optimization method based on crop yield prediction in saline-alkali areas according to claim 2, characterized in that: In step S2, the soil water and salt transport dynamic prediction is used to construct a spatiotemporal prediction model to simulate the dynamic distribution process of salt and water in the soil. Specifically, based on the standardized fusion dataset, an adaptive prediction model combining physical correction and spatiotemporal cubic modeling is used to predict the soil water and salt transport dynamics, obtaining water and salt transport dynamic data. This includes the following steps: Step S21: Data resampling, specifically, by unifying the spatial resolution and time step, interpolating, repairing missing values ​​and rasterizing the original soil water and salt data, and reconstructing it through spatiotemporal standardization to obtain a standardized tensor sequence; Step S22: Spatiotemporal cube structure modeling, specifically by encapsulating the spatial raster tensor of continuous time steps into spatiotemporal cube input units, constructing a cube data structure, and obtaining the spatiotemporal input tensor; Step S23: Spatiotemporal feature extraction and encoding, specifically, performing convolution calculations on the spatiotemporal cube through a three-dimensional convolutional network to encode spatiotemporal features and obtain latent spatial state features; Step S24: Deconvolutional Long Short-Term Memory Network Decoding, specifically by inputting the latent spatial state features into the convolutional long short-term memory network and combining it with the deconvolutional layer for decoding, performing temporal prediction and recovery, and obtaining the water and salt state prediction frame for the future time step; Step S25: Construct a physical constraint loss function. Specifically, by introducing the water-salt diffusion equation and conservation relationship, a regularization constraint based on mass conservation and diffusion gradient is constructed to correct the prediction error and obtain the physical consistency loss. Step S26: Prediction of soil water and salt transport dynamics, specifically by jointly minimizing the mean square error loss and physical consistency loss, to train and infer the network model and obtain water and salt transport dynamic data.

4. The planting optimization method based on crop yield prediction in saline-alkali areas according to claim 3, characterized in that: In step S31, the input factor extraction specifically involves extracting the target crop yield value as a response variable, and extracting soil salinity, moisture content, air temperature, crop type code, and sowing time to form an input matrix. Variables and factors are then constructed to obtain the original input factor matrix and the observed response variable. In step S32, the salt stress threshold interval division is specifically performed by dividing the soil salinity level into multiple threshold segments using a variation point detection algorithm, and then performing salinity interval segmentation processing to obtain a salinity threshold set. In step S33, the physiological response factor function is modeled, specifically by constructing a nonlinear response function based on exponential decay and baseline correction within each salinity interval, performing interval fitting calculations, and obtaining a piecewise salinity response function.

5. The planting optimization method based on crop yield prediction in saline-alkali areas according to claim 4, characterized in that: In step S34, the covariate coupling response correction specifically involves constructing weighted modulation factors for moisture content, air temperature, sowing time, and crop type to correct the salinity response function, thereby obtaining a covariate coupling correction model. In step S35, the parameter optimization model training specifically involves minimizing the yield prediction error loss function and combining it with regularization constraints to perform parameter optimization training, thereby obtaining the final crop salt tolerance response model.

6. The planting optimization method based on crop yield prediction in saline-alkali areas according to claim 5, characterized in that: In step S4, the coupled modeling of salt transport tolerance is used to integrate soil water and salt changes with crop salt tolerance models to simulate the response characteristics of crops in a dynamic salt environment. Specifically, based on the water and salt transport dynamic data and the crop salt tolerance response model, a time-series coupled modeling method combining dynamic water and salt simulation with crop response regulation mechanisms is used to perform coupled modeling of salt transport tolerance to obtain a coupled prediction model, including the following steps: Step S41: Construction of dynamic water and salt state sequence, specifically by extracting salt and water data during the crop growth period, constructing time-step driving variables, and obtaining water and salt driving matrix; Step S42: Design of physiological response memory factors, specifically by constructing a salt stress accumulation index and using an exponential decay weighting function to model the crop salt stress memory effect and obtain time step memory factors; Step S43: Design of stage-sensitive regulatory factors, specifically by identifying different growth stages of crops and setting stage sensitivity coefficients, and then modifying the salt sensitivity to obtain stage-sensitive regulatory factors. Step S44: Dynamic coupling of crop response model, specifically by calling the crop salt tolerance response model, combining time step memory factor and stage regulation factor, to perform dynamic yield calculation and obtain dynamic response yield sequence; Step S45: Coupled prediction model construction, specifically, by calibrating and optimizing the dynamic response yield sequence based on the principle of minimizing yield prediction error, to obtain the final coupled prediction model.

7. The planting optimization method based on crop yield prediction in saline-alkali areas according to claim 6, characterized in that: In step S5, the crop yield prediction and planting optimization are used to calculate the expected yield of crops planted in a specific area based on the output of the coupled model and environmental parameters. Specifically, the coupled prediction model is used to predict crop yield to obtain regional yield prediction data. By analyzing the spatial distribution of the regional yield prediction data, high-yield and low-yield difference areas are identified. Combined with the salt tolerance level and yield threshold of different crop varieties, variety selection and matching are performed. Then, based on the spatial distribution characteristics and variety matching results, a sowing area configuration plan is generated. Combined with water and salt transport dynamic data, field management measures such as irrigation, fertilization and salt removal are scheduled and optimized to comprehensively output regional planting optimization strategies and obtain planting optimization data.

8. A planting optimization system based on crop yield prediction in saline-alkali areas, used to implement the planting optimization method based on crop yield prediction in saline-alkali areas as described in any one of claims 1-7, characterized in that: It includes a data acquisition and fusion processing module, a soil water and salt transport dynamic prediction module, a crop salt tolerance modeling module, a transport and salt tolerance coupled modeling module, and a planting optimization module.

9. The planting optimization system based on crop yield prediction in saline-alkali areas according to claim 8, characterized in that: The data acquisition and fusion processing module is used for data acquisition and fusion processing. Through data acquisition and fusion processing, a standardized fusion dataset is obtained, and the standardized fusion dataset is sent to the soil water and salt transport dynamic prediction module and the crop salt tolerance modeling module. The soil water and salt transport dynamic prediction module is used to predict soil water and salt transport dynamics. Through the prediction of soil water and salt transport dynamics, water and salt transport dynamic data is obtained, and the water and salt transport dynamic data is sent to the transport salt tolerance coupling modeling module. The crop salt tolerance modeling module is used for crop salt tolerance modeling. Through crop salt tolerance modeling, a crop salt tolerance response model is obtained, and the crop salt tolerance response model is sent to the transport salt tolerance coupling modeling module. The transport salt tolerance coupling modeling module is used for transport salt tolerance coupling modeling. Through transport salt tolerance coupling modeling, a coupling prediction model is obtained, and the coupling prediction model is sent to the planting optimization module. The planting optimization module is used for crop yield prediction and planting optimization. Through crop yield prediction and planting optimization, regional yield prediction data and planting optimization strategies are obtained.

Images