Crop straw returning soil fertility evaluation method based on Internet of Things

Abstract

The invention discloses a crop straw returning soil fertility evaluation method based on the Internet of Things, particularly relates to the technical field of agricultural soil evaluation, and is used for solving the problem that the risk that the negative effect of a soil physical structure exceeds the positive effect cannot be warned in time in the prior art. A straw returning area is divided into a plurality of sub-areas, straw returning uniformity and straw processing quality are analyzed based on the Internet of Things, and soil risk sub-areas are determined. And then, analyzing the time sequence data by adopting a long-short-term memory network model, predicting the probability that the negative effect exceeds the positive effect, evaluating the spatial effect of the adjacent sub-regions through a geographically weighted regression model, and finally comprehensively evaluating and early warning the out-of-control degree of the straw returning to the soil fertility. The dynamic monitoring and evaluation precision of the soil structure change is improved, and the problems of soil compaction, fertility decline and the like are effectively prevented.

Description

Technical Field

[0001] The present invention relates to the technical field of agricultural soil assessment, and more specifically, to a method for assessing soil fertility by returning crop straw to fields based on the Internet of Things. Background Art

[0002] Returning crop straw to the field is an important farmland management measure. It can not only increase the organic matter content in the soil, improve the soil structure, and enhance the soil's ability to retain water and fertilizer, but also reduce the environmental pollution caused by straw burning. Returning straw to the field promotes the decomposition of straw through physical or biological means, provides nutrients for crop growth, and enhances soil fertility.

[0003] Among them, the dual effects of returning straw to the field on the physical structure of the soil include positive effects and negative effects. The positive effect is mainly reflected in the increase of organic matter content in the soil after straw decomposition, the improvement of aggregate structure, the enhancement of soil water retention and air permeability, and the improvement of soil fertility. The negative effect is manifested in the case of uneven return to the field or poor quality of straw crushing. Straw accumulation may cause soil pore blockage, reduce air permeability and permeability, form soil compaction or hardening, and hinder the growth of crop roots.

[0004] The existing technology lacks dynamic and accurate assessment of soil structure changes in straw return areas, and is unable to promptly warn of the risk that the negative effects of soil physical structure will outweigh the positive effects, thereby affecting soil fertility and hindering crop growth.

[0005] In order to solve the above problems, a technical solution is now provided. Summary of the invention

[0006] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a method for evaluating soil fertility by returning crop straw to fields based on the Internet of Things to solve the problems raised in the above-mentioned background technology.

[0007] To achieve the above object, the present invention provides the following technical solutions: A method for evaluating soil fertility by returning crop straw to fields based on the Internet of Things comprises the following steps: The straw return area is divided into multiple sub-areas, and the uniformity of straw return and straw treatment quality are analyzed based on the Internet of Things to determine the soil risk sub-areas; By analyzing the time series data of soil risk sub-regions, a dual-effect time series model was established using a long short-term memory network model to predict the probability that negative effects exceed positive effects in the time dimension. The dual effects of soil risk sub-regions and their adjacent sub-regions were spatially analyzed using a geographically weighted regression model to assess the potential risk of negative effects in soil risk sub-regions in the spatial dimension. Based on the probability of negative effects exceeding positive effects in the time dimension and the potential risk of negative effects in the soil risk sub-regions in the space dimension, the negative effects of returning straw to the fields on the soil physical structure in the soil risk sub-regions are evaluated, and the degree of uncontrollable soil fertility caused by returning straw to the fields in the soil risk sub-regions is warned based on the evaluation results. The dual effects of soil risk sub-regions and their adjacent sub-regions were spatially analyzed using a geographically weighted regression model to assess the potential risk of negative effects in soil risk sub-regions in the spatial dimension. Specifically: Collect the geographical coordinates, soil physical structure parameters and straw return-related data of each soil risk sub-area and its adjacent sub-areas to form a spatial data set; The spatial weight function is used to analyze the impact of the negative effect data of adjacent sub-regions on the soil risk sub-region; The negative effects of the soil risk sub-regions themselves were analyzed through the geographically weighted regression model, and the contribution of the soil physical structure changes in the soil risk sub-regions to the negative effects was calculated; The Markov Chain Monte Carlo algorithm is used to simulate the feedback effect of the adjacent sub-region on its own negative effects, and the possibility of the negative effect of the adjacent sub-region on the soil risk sub-region is calculated; The contribution of soil physical structure changes in soil risk sub-regions to negative effects and the possibility of negative effects of adjacent sub-regions on soil risk sub-regions are comprehensively analyzed, and the spatial negative hidden danger index is calculated to evaluate the hidden danger degree of negative effects in soil risk sub-regions in the spatial dimension. Among them, the expression of spatial negative hidden danger index is: ;in, is the spatial negative risk index, represents the possibility of negative effects of adjacent sub-regions on soil risk sub-regions, is the total number of adjacent sub-regions, and represent the negative effect value of the soil risk sub-region and the weight factor of the possibility of the negative effect of the adjacent sub-region on the soil risk sub-region, and and Both are greater than 0.

[0008] In a preferred embodiment, the straw return area is divided into multiple sub-areas, and the uniformity of straw return and straw treatment quality are analyzed based on the Internet of Things to determine the soil risk sub-areas, specifically: Based on soil type, topography and historical data, the straw return area is divided into multiple sub-areas and corresponding IoT devices are set up; Use IoT devices to collect uniformity data of straw return to fields in each sub-region, including the distribution density and coverage rate of straw in each sub-region; Collect straw processing quality data of each sub-area, including the particle size and length of the crushed straw; Each sub-area is analyzed based on the straw return uniformity data and straw treatment quality data. According to the analysis results, the sub-areas with potential risks in soil structure are determined and marked as soil risk sub-areas.

[0009] In a preferred embodiment, by analyzing the time series data of the soil risk sub-areas, a long short-term memory network model is used to establish a dual-effect time series model to predict the probability of negative effects exceeding positive effects in the time dimension, specifically: Collect time series data of soil risk sub-areas, including soil physical structure data and straw return-related data, to form a continuous time series data set at predetermined time intervals; Preprocess the time series data, including data cleaning, missing value filling and normalization; Construct a long short-term memory network model and set the number of layers, nodes and learning rate of the model according to the characteristics of the time series data of the soil risk sub-regions; The preprocessed time series data were used to train the long short-term memory network model, and the model parameters were adjusted through multiple iterations to generate a dual-effect time series model for the soil risk sub-regions. The probability of negative effects exceeding positive effects is predicted based on the trained model, and the probability of negative effects exceeding positive effects in the soil risk sub-area in the time dimension is calculated through the output results of the model.

[0010] In a preferred embodiment, the probability that the negative effect of the soil risk sub-area exceeds the positive effect in the time dimension is calculated, and its expression is: ;in, is the probability that the negative effect exceeds the positive effect in the time dimension, It's time. is the linear combination value output by the LSTM model, It is the base of the natural logarithm of the mathematical constant; The linear combination value output by the LSTM model is: ;in, is the weight matrix, is the input vector of time series data, is the bias vector.

[0011] In a preferred embodiment, the impact intensity of the negative effect data of the adjacent sub-regions on the soil risk sub-region is analyzed by a spatial weight function, specifically: Calculate the distance between the adjacent sub-region and the soil risk sub-region, and assign different weights to each adjacent sub-region; The weight function can be in the form of a Gaussian kernel function: ;in, is the weight function, is the scale parameter of the influence range, is the spatial distance, is the number of the sub-area adjacent to the soil risk sub-area; Analyze the impact of the negative effects of adjacent sub-regions on the target region and obtain the impact strength of each sub-region: ;in, is the influence strength of the sub-region, is the negative effect data of the adjacent sub-region.

[0012] In a preferred embodiment, the negative effects of the soil risk sub-region itself are analyzed by a geographically weighted regression model, and the contribution of the soil physical structure changes in the soil risk sub-region to the negative effects is calculated, specifically: Regression analysis was performed using the following formula: ;in, is the negative effect value of the soil risk sub-area, is the input soil physical structure parameter, is the constant term of the model, is the regression coefficient, is the sum of soil physical structure parameters.

[0013] In a preferred embodiment, the Markov Chain Monte Carlo algorithm is used to simulate the feedback effect of the adjacent sub-region on its own negative effect, and the possibility of the negative effect of the adjacent sub-region on the soil risk sub-region is calculated, specifically: The initial state of each adjacent sub-region is obtained, and based on the initial state, the transition probability of its subsequent state is calculated. The transition probability of the subsequent state is determined by the Markov chain model, and its expression is: ;in, is the transition probability to the subsequent state, is the functional form of state transition, is the current state in the Markov chain, is a set of parameters, represents the next state in the Markov chain, is the number of the state; Through Monte Carlo random sampling, the negative effects of adjacent sub-regions are simulated multiple times, and the state transition probability of the adjacent sub-region transmitting the negative effect to the soil risk sub-region at different time steps is calculated, thereby obtaining the possibility of the negative effect of the adjacent sub-region on the soil risk sub-region.

[0014] In a preferred embodiment, the negative effects of returning straw to the soil in the soil risk sub-regions are evaluated based on the probability of negative effects exceeding positive effects in the time dimension and the potential risk of negative effects in the soil risk sub-regions in the space dimension, specifically: Set a threshold for the probability that negative effects exceed positive effects in the time dimension; set a threshold for the spatial negative hidden danger index; If the probability of negative effects exceeding positive effects in the current time dimension is greater than its corresponding threshold, and the current spatial negative hidden danger index is greater than its corresponding threshold, then the soil risk sub-area is assessed as high risk; If the probability of negative effects exceeding positive effects in the current time dimension is greater than the corresponding threshold, but the current spatial negative hidden danger index is less than or equal to the corresponding threshold, or if the probability of negative effects exceeding positive effects in the current time dimension is less than or equal to the corresponding threshold, but the current spatial negative hidden danger index is greater than the corresponding threshold, then the soil risk sub-area is assessed as medium risk; If the probability of negative effects exceeding positive effects in the current time dimension is less than or equal to its corresponding threshold, and the current spatial negative hidden danger index is less than or equal to its corresponding threshold, then the soil risk sub-area is assessed as low risk; The degree of uncontrolled soil fertility caused by returning straw to the field in the early warning soil risk sub-area is as follows: For soil risk sub-areas assessed as medium risk, a medium-level warning will be issued; For soil risk sub-areas assessed as high risk, a high priority early warning is issued.

[0015] The technical effects and advantages of the method for evaluating soil fertility by returning crop straw to fields based on the Internet of Things of the present invention are as follows: 1. By dividing the straw return area into multiple sub-areas and analyzing the uniformity of straw return and straw processing quality based on the Internet of Things, the soil risk sub-areas can be accurately and dynamically determined. The identification of soil risk sub-areas can avoid overly extensive and inaccurate assessments of the entire area. This division and analysis mechanism improves the efficiency of monitoring changes in soil physical structure, allowing the negative effects that may occur during the straw return process to be identified and evaluated early, thereby providing a more accurate soil management plan for crop growth.

[0016] 2. By combining the long-short-term memory network model and the geographically weighted regression model, not only can the time point when the negative effect exceeds the positive effect be predicted through time series data, but also the impact of adjacent areas on the target area can be analyzed through spatial data, and the degree of hidden dangers of negative effects in the soil risk sub-area can be accurately assessed. Based on this, it is possible to warn of the risk of soil fertility getting out of control in real time, so that managers can take timely measures to adjust the straw return strategy, avoid fertility loss due to soil structure deterioration, and ensure the stability of crop growth and the sustainability of soil ecology. This systematic dynamic assessment method has greatly improved the accuracy of soil management and reduced the long-term negative impact of improper straw handling on soil fertility. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 This is a schematic diagram of a method for evaluating soil fertility by returning crop straw to fields based on the Internet of Things according to the present invention. DETAILED DESCRIPTION

[0018] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0019] Example Figure 1 The present invention provides a method for evaluating soil fertility by returning crop straw to fields based on the Internet of Things, which comprises the following steps: The straw return area is divided into multiple sub-areas, and the uniformity of straw return and the quality of straw treatment are analyzed based on the Internet of Things to determine the soil risk sub-areas.

[0020] By analyzing the time series data of soil risk sub-areas, a dual-effect time series model was established using the long short-term memory network model to predict the probability that the negative effect would exceed the positive effect in the time dimension.

[0021] The dual effects of soil risk sub-regions and their adjacent sub-regions were spatially analyzed through the geographically weighted regression model to assess the potential risk of negative effects in soil risk sub-regions in the spatial dimension.

[0022] Based on the probability that negative effects exceed positive effects in the time dimension and the hidden danger degree of negative effects occurring in the soil risk sub-area in the spatial dimension, the negative effects of straw returning to the fields on the soil physical structure in the soil risk sub-area are evaluated, and the degree of uncontrollable soil fertility caused by straw returning to the fields in the soil risk sub-area is warned based on the evaluation results.

[0023] The straw return area is divided into multiple sub-areas. The uniformity of straw return and straw treatment quality are analyzed based on the Internet of Things to determine the soil risk sub-areas, which are as follows: Based on soil type, topography and historical data, the straw return area is divided into multiple sub-areas and corresponding IoT devices are set up: The straw return area was divided into multiple sub-areas based on soil type, topography and historical data. Soil types can include sand, loam and clay, while topography is based on elevation, slope and drainage.

[0024] Each sub-region was divided based on its homogeneity to ensure the accuracy of data collection.

[0025] IoT devices, including sensors and data acquisition terminals, are installed in each sub-area. The sensors can collect soil moisture, temperature, coverage and other parameters in real time, and transmit the data to the centralized control platform through wireless networks.

[0026] IoT devices are used to collect uniformity data of straw return to fields in each sub-region. The uniformity data of straw return to fields includes the distribution density and coverage rate of straw in each sub-region: Optical sensors or laser scanners in IoT devices are used to collect data on the uniformity of straw return to the field in each sub-area.

[0027] Through laser scanning image analysis, the straw coverage per square meter is calculated, and the sparse or dense distribution is used as one of the main evaluation indicators.

[0028] Based on the real-time feedback data from the optical sensor, the percentage of the surface covered by straw in each sub-area is measured and recorded in the system.

[0029] Collect straw processing quality data for each sub-area, including the particle size and length of the crushed straw: The particle size sensor detects the crushed particle size of the straw and records the average size of the straw particles in each sub-area.

[0030] The straw cutting monitoring device measures the average length of the crushed straw and analyses its length distribution to ensure that it meets the farmland return standards (e.g. the length should be between 5-10 cm).

[0031] Each sub-area is analyzed based on the straw return uniformity data and straw treatment quality data. Sub-areas with potential risks in soil structure are identified based on the analysis results and marked as soil risk sub-areas: Each sub-area was analyzed based on the collected data on straw distribution density, coverage, crushing particle size and length.

[0032] Compare each data with the preset standard range: the standard range of distribution density, coverage, crushing particle size and length is set by historical data or experience, such as: Distribution density: There should be at least a certain unit amount of straw per square meter of the straw-covered area.

[0033] Coverage rate: The straw coverage rate of each sub-area should reach a certain proportion (for example, more than 80%).

[0034] Particle size: The particle size should be within the specified range.

[0035] Crushing length: The straw length should be within the specified range (such as 5-10 cm).

[0036] Use IoT systems to automatically compare actual data with preset standard ranges: As long as any one of the straw distribution density, coverage rate, particle size and length does not meet the preset standard range, it will be marked as a soil risk sub-area.

[0037] By analyzing the time series data of soil risk sub-areas, a dual-effect time series model was established using the long short-term memory network model to predict the probability of negative effects exceeding positive effects in the time dimension, specifically: Collect time series data of soil risk sub-areas. The time series data includes soil physical structure data and straw return data. A continuous time series data set is formed at predetermined time intervals: Multiple sensor networks are deployed in the soil risk sub-areas to ensure that accurate and complete time series data can be continuously collected. The collected data includes the following: Soil physical structure parameters: including porosity (i.e. the ratio of pores inside the soil), soil compaction (soil compaction caused by moisture content and mechanical pressure), temperature and humidity. These data can be obtained through moisture sensors, temperature sensors and density sensors installed in the soil, and the frequency is collected once every hour.

[0038] Data related to straw return to the field: including straw decomposition rate, microbial activity and uniformity of straw distribution. The decomposition rate is determined by analyzing the changes in the content of cellulose, hemicellulose and lignin in the straw. These data are obtained through IoT sensors, which are installed in multiple locations in the sub-region to ensure coverage of representative areas of each sub-region.

[0039] Preprocess the time series data, including data cleaning, missing value filling and normalization: Data cleaning: Cleans noise and outliers due to sensor errors or environmental interference. For example, extreme temperature or humidity values ​​may be caused by equipment failure, and such data will be identified and deleted.

[0040] Missing value filling: Due to network transmission or equipment failure, some data points may be missing. To prevent these missing data from affecting model training, linear interpolation is used to fill the missing values.

[0041] Normalization: In order to ensure the dimensional consistency between different physical quantities, all data need to be normalized.

[0042] Construct a long short-term memory network model and set the number of layers, nodes and learning rate of the model according to the characteristics of the time series data of the soil risk sub-areas: The Long Short-Term Memory (LSTM) model is a recurrent neural network that specifically processes time series data and can effectively capture the temporal dependencies in the data. The model construction process is as follows: Input layer: The number of nodes in the input layer is consistent with the number of features of the time series data. Assuming that there are multiple features at each time point, including soil porosity, compaction, humidity, temperature, and straw decomposition rate, the number of nodes in the input layer is the number of features.

[0043] Hidden layer: The core of the LSTM model is the LSTM units in the hidden layer, which can effectively capture time dependencies through the "memory gating mechanism". In this embodiment, the hidden layer is set to two layers, the first layer contains 64 LSTM units, and the second layer contains 32 LSTM units.

[0044] Output layer: The number of nodes in the output layer is 1, which is used to generate the prediction result, that is, the probability that the negative effect exceeds the positive effect.

[0045] Learning rate: The learning rate determines the speed at which the model parameters are updated. In this embodiment, the learning rate is set to 0.001 to ensure stable convergence of the model.

[0046] The loss function of the LSTM model uses the mean square error (MSE), which is used to measure the gap between the predicted value and the actual value. Specifically, the mean square error is the square of the difference between the actual value of each sample and the model's predicted value, and then the average value of all samples is calculated.

[0047] The preprocessed time series data is used to train the long short-term memory network model, and the model parameters are adjusted through multiple iterations to generate a dual-effect time series model for the soil risk sub-region: Time series data is input into the model in chronological order, and the model continuously adjusts the internal weight parameters through the back propagation algorithm and gradient descent method. The key steps of the training process are as follows: The data is divided into a training set and a validation set, and the cross-validation method is used for training. At each iteration, the model adjusts its internal parameters so that the loss function gradually decreases until the model reaches a stable convergence state.

[0048] Through multiple iterations, the model parameters will be gradually optimized, and the training process adopts a dynamic learning rate adjustment mechanism to accelerate the convergence speed.

[0049] During training, the model uses the input time series data and combines the positive and negative effects in the historical data to learn the changing patterns of the dual effects in the soil risk sub-areas.

[0050] Based on the trained model, the probability of negative effects exceeding positive effects is predicted. Through the output of the model, the probability of negative effects exceeding positive effects in the soil risk sub-area in the time dimension is calculated: The probability that the negative effect of the soil risk sub-area exceeds the positive effect in the time dimension is calculated as follows: ;in, is the probability that the negative effect exceeds the positive effect in the time dimension, It's time. is the linear combination value output by the LSTM model, It is the base of the mathematical constant natural logarithm.

[0051] Among them, the linear combination value output by the LSTM model is: ;in, is the weight matrix, is the input vector of time series data, is the bias vector.

[0052] is a probability value between 0 and 1. When it is close to 1, it means that at time There is a high probability that the negative effects outweigh the positive effects.

[0053] Right now The larger it is, the greater the probability that the negative effects will exceed the positive effects in the soil risk sub-area in the time dimension.

[0054] The dual effects of soil risk sub-regions and their adjacent sub-regions were spatially analyzed using a geographically weighted regression model to assess the potential risk of negative effects in soil risk sub-regions in the spatial dimension. Specifically: The geographical coordinates, soil physical structure parameters and straw return-related data of each soil risk sub-area and its adjacent sub-areas were collected to form a spatial dataset: Geographic coordinates: The spatial location of each sub-region, including the longitude and latitude of its boundary points.

[0055] Soil physical structure parameters: including porosity, compaction, moisture, temperature, etc. Porosity and compaction are monitored in real time through installed IoT sensors and recorded at set time intervals. The data collection frequency can be set to once an hour.

[0056] Data related to straw return to the field: including straw decomposition rate, straw cover density and microbial activity. The straw decomposition rate is calculated by monitoring the reduction rate of the main organic components in the straw (such as cellulose, hemicellulose and lignin). The decomposition rate unit can be the degradation percentage per hour.

[0057] Geographic coordinates, soil physical structure parameters and data related to straw return to fields are automatically collected and summarized through the sensor network to form a complete spatial data set.

[0058] The spatial weight function is used to analyze the impact of the negative effect data of adjacent sub-regions on the soil risk sub-region: A spatial weight function was constructed to assign different weights to each adjacent sub-region by calculating the distance between the adjacent sub-region and the soil risk sub-region.

[0059] Assume that the geographic coordinates of each adjacent sub-region are , the geographic center coordinates of the soil risk sub-area are . Define the spatial distance as: ;in, is the spatial distance, is the number of the sub-area adjacent to the soil risk sub-area.

[0060] The weight function can be in the form of a Gaussian kernel function: ;in, is the weight function, is the scale parameter of the influence range.

[0061] Among them, the scale parameter of the influence range is determined by fitting the experimental data to ensure that the influence weights of adjacent sub-areas are reasonably attenuated according to the distance.

[0062] Through the weight function, the impact of the negative effects of adjacent sub-regions on the target region is analyzed to obtain the impact strength of each sub-region. The impact strength of the sub-region is calculated by the following formula: ;in, is the influence strength of the sub-region, It is the negative effect data of the adjacent sub-area (such as the soil compaction change rate or porosity decrease rate).

[0063] The negative effects of the soil risk sub-region itself were analyzed through the geographical weighted regression model, and the contribution of soil physical structure changes in the soil risk sub-region to the negative effects was calculated: The geographically weighted regression model is a location-based regression model that can analyze the relationship between spatial data.

[0064] Assuming that the negative effect of a point within the soil risk sub-area is affected by its spatial coordinates and the changes in its soil physical structure, regression analysis is performed using the following formula: ;in, is the negative effect value of the soil risk sub-area; are input soil physical structure parameters (e.g. porosity, compaction, etc.), representing different factors that influence the negative effects of soil risk sub-areas; is the constant term of the model, also called the intercept; is the regression coefficient, which represents the weight of each input variable’s negative effect on the soil risk sub-region, obtained by fitting the spatial data through the geographically weighted regression model; is the sum of soil physical structure parameters.

[0065] Specifically, the negative effect value of the soil risk sub-area indicates the negative impact caused by changes in the soil physical structure in the target area. This negative effect may include soil compaction, decreased porosity, decreased soil permeability, and other changes that are not conducive to crop growth.

[0066] The larger the negative effect value of a soil risk sub-area, the greater the contribution of soil physical structure changes in the soil risk sub-area to the negative effect.

[0067] The Markov Chain Monte Carlo algorithm is used to simulate the feedback effect of the adjacent sub-region on its own negative effects, and the possibility of the negative effect of the adjacent sub-region on the soil risk sub-region is calculated: The Markov Chain Monte Carlo algorithm is a random sampling algorithm based on probability distribution and is suitable for dealing with complex spatial effects.

[0068] The initial state of each adjacent sub-region is obtained, and the initial state represents the initial value of its negative effect. Based on the initial state, the transition probability of its subsequent state is calculated. The transition probability of the subsequent state is determined by the Markov chain model, and its expression is: ;in, is the transition probability of the subsequent state; is the functional form of state transition, expressed by and The functional form of the determined transition probabilities; is the current state in the Markov chain; Is a collection of parameters that affect the state from Transfer to All factors or parameters determine the rules for state transition; represents the next state in the Markov chain; is the number of the state.

[0069] Through Monte Carlo random sampling, the negative effects of adjacent sub-regions were simulated multiple times, and the state transition probability of the adjacent sub-region transmitting the negative effect to the soil risk sub-region at different time steps was calculated, and finally the possibility of the negative effect of the adjacent sub-region on the soil risk sub-region was obtained.

[0070] The contribution of soil physical structure changes in soil risk sub-regions to negative effects and the possibility of negative effects of adjacent sub-regions on soil risk sub-regions are comprehensively analyzed to assess the potential risk of negative effects in soil risk sub-regions in the spatial dimension: Calculate the spatial negative hidden danger index, the expression is: ;in, is the spatial negative risk index, represents the possibility of negative effects of adjacent sub-regions on soil risk sub-regions, is the total number of adjacent sub-regions, and represent the negative effect value of the soil risk sub-region and the weight factor of the possibility of the negative effect of the adjacent sub-region on the soil risk sub-region, and and Both are greater than 0.

[0071] The larger the spatial negative hidden danger index is, the higher the hidden danger level of negative effects occurring in the soil risk sub-area in the spatial dimension.

[0072] Based on the probability of negative effects exceeding positive effects in the time dimension and the potential risk of negative effects in the soil risk sub-area in the space dimension, the negative effects of returning straw to the soil in the soil risk sub-area on the soil physical structure were evaluated, specifically: Thresholds are set for the probability of negative effects exceeding positive effects in the temporal dimension and the spatial negative hazard index to ensure that the potential negative effects of soil risk sub-areas can be reasonably assessed.

[0073] Set the threshold of the probability of negative effects exceeding positive effects in the time dimension: Analyze the moments when negative effects exceeded positive effects in the soil risk sub-area in the past time period based on historical data. By counting the probability values ​​of these time points, find the key probability range that leads to negative effects. Use statistical analysis methods, such as selecting the 75th percentile of the time probability of negative effects in historical data as the threshold. This threshold is used to determine the extent to which negative effects may exceed positive effects in the current time period.

[0074] Set the threshold of the spatial negative hidden danger index: Based on historical spatial data, analyze the spatial negative hidden danger index of the soil risk sub-area when negative effects occur. Count the spatial index in these events and find the key hidden danger index interval. Through data analysis and regression model fitting, determine the critical value of the hidden danger index, such as the 75th percentile, as a baseline for evaluating the hidden dangers of negative effects in the spatial dimension.

[0075] Compare the probability of negative effects exceeding positive effects in the current time dimension with their corresponding thresholds: If the probability that the negative effect exceeds the positive effect in the current time dimension is greater than the corresponding threshold, it is judged that the probability of the negative effect caused by returning straw to the field in the current time period is high.

[0076] Compare the current spatial negative risk index with its corresponding threshold: If the current spatial negative hazard index is greater than its corresponding threshold, it is judged that the risk of negative effects in the soil risk sub-area in the spatial dimension is high.

[0077] If the probability that the negative effect exceeds the positive effect in the current time dimension is greater than its corresponding threshold, and the current spatial negative hidden danger index is greater than its corresponding threshold, the soil risk sub-area is assessed as high risk; this situation means that the possibility of negative effects in the time dimension is high, and there are also significant hidden dangers in the spatial dimension, so the area is in a high-risk state.

[0078] If the probability that the negative effect exceeds the positive effect in the current time dimension is greater than its corresponding threshold, but the current spatial negative hidden danger index is less than or equal to its corresponding threshold, or if the probability that the negative effect exceeds the positive effect in the current time dimension is less than or equal to its corresponding threshold, but the current spatial negative hidden danger index is greater than its corresponding threshold, then the soil risk sub-area is assessed as medium risk; this situation indicates that there is a higher risk of negative effects in one dimension, while the risk in another dimension is smaller, so the comprehensive assessment is medium risk.

[0079] If the probability of negative effects exceeding positive effects in the current time dimension is less than or equal to its corresponding threshold, and the current spatial negative hidden danger index is less than or equal to its corresponding threshold, then the soil risk sub-area is assessed to be low risk; in this case, both the time and space dimensions show low risk, indicating that the probability of negative effects occurring in the area in the short term is very small and is in a relatively stable state.

[0080] The degree of uncontrolled soil fertility caused by returning straw to the field in the early warning soil risk sub-area is as follows: The degree of loss of control refers to the possibility that the soil physical structure in the soil risk sub-area will deteriorate due to returning straw to the field, which will have a serious negative impact on soil fertility. The higher the degree of loss of control, the more difficult it will be to restore the soil fertility in the area through conventional measures, which may lead to reduced crop yields, soil deterioration and even ecological imbalance.

[0081] When the soil risk sub-area is assessed as medium risk, it indicates that the negative effects of soil physical structure have begun to appear, and the impact on soil fertility has gradually intensified. If intervention measures are not taken in time, soil fertility may become out of control. When the soil risk sub-area is assessed as high risk, it indicates that the serious deterioration of soil physical structure has had a significant negative effect on soil fertility, and the degree of soil fertility in the area is out of control, and it is extremely difficult to recover. Emergency measures must be taken immediately.

[0082] For soil risk sub-areas assessed as medium risk, a medium-level warning is issued, and it is recommended to regularly monitor the soil conditions in the area and gradually take measures to improve the way of returning straw to the fields to prevent further deterioration.

[0083] For soil risk sub-areas assessed as high-risk, high-priority warnings are issued and emergency measures are immediately initiated, such as reducing the amount of straw returned to the fields and adjusting fertilizer use to prevent soil fertility from further getting out of control.

[0084] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters and thresholds in the formula are set by technicians in this field according to actual conditions.

[0085] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium, or may be transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from one website, computer, server or data center to another website, computer, server or data center by wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media sets. The available medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium may be a solid-state hard disk.

[0086] Those of ordinary skill in the art will appreciate that the modules and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0087] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and modules described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0088] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the modules is only a logical function division. There may be other division methods in actual implementation, such as multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or modules, which can be electrical, mechanical or other forms.

[0089] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical modules, and may be located in one place or distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0090] In addition, each functional module in each embodiment of the present application may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0091] If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage media include: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, and other media that can store program codes.

[0092] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0093] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for evaluating soil fertility by returning crop straw to fields based on the Internet of Things, characterized in that: The steps include: The straw return area is divided into multiple sub-areas, and the uniformity of straw return and straw treatment quality are analyzed based on the Internet of Things to determine the soil risk sub-areas; By analyzing the time series data of soil risk sub-regions, a dual-effect time series model was established using a long short-term memory network model to predict the probability that negative effects exceed positive effects in the time dimension. The dual effects of soil risk sub-regions and their adjacent sub-regions were spatially analyzed using a geographically weighted regression model to assess the potential risk of negative effects in soil risk sub-regions in the spatial dimension. Based on the probability of negative effects exceeding positive effects in the time dimension and the potential risk of negative effects in the soil risk sub-regions in the space dimension, the negative effects of returning straw to the fields on the soil physical structure in the soil risk sub-regions are evaluated, and the degree of uncontrollable soil fertility caused by returning straw to the fields in the soil risk sub-regions is warned based on the evaluation results. The dual effects of soil risk sub-regions and their adjacent sub-regions were spatially analyzed using a geographically weighted regression model to assess the potential risk of negative effects in soil risk sub-regions in the spatial dimension. Specifically: Collect the geographical coordinates, soil physical structure parameters and straw return-related data of each soil risk sub-area and its adjacent sub-areas to form a spatial data set; The spatial weight function is used to analyze the impact of the negative effect data of adjacent sub-regions on the soil risk sub-region; The negative effects of the soil risk sub-regions themselves were analyzed through the geographically weighted regression model, and the contribution of the soil physical structure changes in the soil risk sub-regions to the negative effects was calculated; The Markov Chain Monte Carlo algorithm is used to simulate the feedback effect of the adjacent sub-region on its own negative effects, and the possibility of the negative effect of the adjacent sub-region on the soil risk sub-region is calculated; The contribution of soil physical structure changes in soil risk sub-regions to negative effects and the possibility of negative effects of adjacent sub-regions on soil risk sub-regions are comprehensively analyzed, and the spatial negative hidden danger index is calculated to evaluate the hidden danger degree of negative effects in soil risk sub-regions in the spatial dimension. Among them, the expression of spatial negative hidden danger index is: ;in, is the spatial negative risk index, represents the possibility of negative effects of adjacent sub-regions on soil risk sub-regions, is the total number of adjacent sub-regions, and represent the negative effect value of the soil risk sub-region and the weight factor of the possibility of the negative effect of the adjacent sub-region on the soil risk sub-region, and and Both are greater than 0.

2. The method for evaluating soil fertility by returning crop straw to fields based on the Internet of Things according to claim 1, characterized in that: The straw return area is divided into multiple sub-areas. The uniformity of straw return and straw treatment quality are analyzed based on the Internet of Things to determine the soil risk sub-areas, which are as follows: Based on soil type, topography and historical data, the straw return area is divided into multiple sub-areas and corresponding IoT devices are set up; Use IoT devices to collect uniformity data of straw return to fields in each sub-region, including the distribution density and coverage rate of straw in each sub-region; Collect straw processing quality data of each sub-area, including the particle size and length of the crushed straw; Each sub-area is analyzed based on the straw return uniformity data and straw treatment quality data. According to the analysis results, the sub-areas with potential risks in soil structure are determined and marked as soil risk sub-areas.

3. The method for evaluating soil fertility by returning crop straw to fields based on the Internet of Things according to claim 1, characterized in that: By analyzing the time series data of soil risk sub-areas, a dual-effect time series model was established using the long short-term memory network model to predict the probability of negative effects exceeding positive effects in the time dimension, specifically: Collect time series data of soil risk sub-areas, including soil physical structure data and straw return-related data, to form a continuous time series data set at predetermined time intervals; Preprocess the time series data, including data cleaning, missing value filling and normalization; Construct a long short-term memory network model and set the number of layers, nodes and learning rate of the model according to the characteristics of the time series data of the soil risk sub-regions; The preprocessed time series data were used to train the long short-term memory network model, and the model parameters were adjusted through multiple iterations to generate a dual-effect time series model for the soil risk sub-regions. The probability of negative effects exceeding positive effects is predicted based on the trained model, and the probability of negative effects exceeding positive effects in the soil risk sub-area in the time dimension is calculated through the output results of the model.

4. The method for evaluating soil fertility by returning crop straw to fields based on the Internet of Things according to claim 3 is characterized in that: The probability that the negative effect of the soil risk sub-area exceeds the positive effect in the time dimension is calculated as follows: ;in, is the probability that the negative effect exceeds the positive effect in the time dimension, It's time. is the linear combination value output by the LSTM model, It is the base of the natural logarithm of the mathematical constant; The linear combination value output by the LSTM model is: ;in, is the weight matrix, is the input vector of time series data, is the bias vector.

5. The method for evaluating soil fertility by returning crop straw to fields based on the Internet of Things according to claim 1, characterized in that: The spatial weight function is used to analyze the impact of the negative effect data of adjacent sub-regions on the soil risk sub-region, specifically: Calculate the distance between the adjacent sub-region and the soil risk sub-region, and assign different weights to each adjacent sub-region; The weight function can be in the form of a Gaussian kernel function: ;in, is the weight function, is the scale parameter of the influence range, is the spatial distance, is the number of the sub-area adjacent to the soil risk sub-area; Analyze the impact of the negative effects of adjacent sub-regions on the target region and obtain the impact strength of each sub-region: ;in, is the influence strength of the sub-region, is the negative effect data of the adjacent sub-region.

6. The method for evaluating soil fertility by returning crop straw to fields based on the Internet of Things according to claim 1, characterized in that: The negative effects of the soil risk sub-region itself were analyzed through the geographical weighted regression model, and the contribution of soil physical structure changes in the soil risk sub-region to the negative effects was calculated, specifically: Regression analysis was performed using the following formula: ;in, is the negative effect value of the soil risk sub-area, is the input soil physical structure parameter, is the constant term of the model, is the regression coefficient, is the sum of soil physical structure parameters.

7. The method for evaluating soil fertility by returning crop straw to fields based on the Internet of Things according to claim 1, characterized in that: The Markov Chain Monte Carlo algorithm is used to simulate the feedback effect of the adjacent sub-region on its own negative effects, and the possibility of the negative effect of the adjacent sub-region on the soil risk sub-region is calculated, specifically: The initial state of each adjacent sub-region is obtained, and based on the initial state, the transition probability of its subsequent state is calculated. The transition probability of the subsequent state is determined by the Markov chain model, and its expression is: ;in, is the transition probability to the subsequent state, is the functional form of state transition, is the current state in the Markov chain, is a set of parameters, represents the next state in the Markov chain, is the number of the state; Through Monte Carlo random sampling, the negative effects of adjacent sub-regions are simulated multiple times, and the state transition probability of the adjacent sub-region transmitting the negative effect to the soil risk sub-region at different time steps is calculated, thereby obtaining the possibility of the negative effect of the adjacent sub-region on the soil risk sub-region.

8. The method for evaluating soil fertility by returning crop straw to fields based on the Internet of Things according to claim 1, characterized in that: Based on the probability of negative effects exceeding positive effects in the time dimension and the potential risk of negative effects in the soil risk sub-area in the space dimension, the negative effects of returning straw to the soil in the soil risk sub-area on the soil physical structure were evaluated, specifically: Set a threshold for the probability that negative effects exceed positive effects in the time dimension; set a threshold for the spatial negative hidden danger index; If the probability of negative effects exceeding positive effects in the current time dimension is greater than its corresponding threshold, and the current spatial negative hidden danger index is greater than its corresponding threshold, then the soil risk sub-area is assessed as high risk; If the probability of negative effects exceeding positive effects in the current time dimension is greater than the corresponding threshold, but the current spatial negative hidden danger index is less than or equal to the corresponding threshold, or if the probability of negative effects exceeding positive effects in the current time dimension is less than or equal to the corresponding threshold, but the current spatial negative hidden danger index is greater than the corresponding threshold, then the soil risk sub-area is assessed as medium risk; If the probability of negative effects exceeding positive effects in the current time dimension is less than or equal to its corresponding threshold, and the current spatial negative hidden danger index is less than or equal to its corresponding threshold, then the soil risk sub-area is assessed as low risk; The degree of uncontrolled soil fertility caused by returning straw to the field in the early warning soil risk sub-area is as follows: For soil risk sub-areas assessed as medium risk, a medium-level warning will be issued; For soil risk sub-areas assessed as high risk, a high priority early warning is issued.

Images